Annotation method, device, and computer-readable storage medium

Abstract

The present application discloses a labeling method, device, and computer-readable storage medium. The method includes: obtaining a first target box of a first target object and a set of target box point positions corresponding to a second target object in a target image, where the set of target box point positions includes multiple initial labeling points of the second target object; in response to a first line segment formed by a first initial labeling point and a second initial labeling point in the set of target box point positions intersecting with the first target box, determining an intersection point of the first line segment and the first target box; determining a target initial labeling point that is within the first target box among the first initial labeling point and the second initial labeling point; replacing the target initial labeling point in the set of target point positions with the intersection point; and determining a second target box of the second target object according to the set of target box point positions. Through the above manner, the present application can reduce the workload of labelers and effectively reduce the labeling difficulty of labelers.

Description

Technical Field

[0001] This application belongs to the technical field of image detection, and particularly relates to a labeling method, device, and computer-readable storage medium. Background Art

[0002] Existing various video algorithms have been very mature, but target detection and tracking for dense crowds are still difficult, mainly because there is a lack of effective and clean data for training models. However, existing labeling methods have problems such as cumbersome labeling and serious occlusion in crowded areas, such as classrooms and streets, and require a lot of effort from labelers. Summary of the Invention

[0003] This application provides a labeling method, device, and computer-readable storage medium, which can reduce the workload of labelers and effectively reduce the labeling difficulty of labelers.

[0004] To solve the above technical problems, in the first aspect of the embodiments of this application, a labeling method is provided. The method includes: obtaining a first target box of a first target object and a set of target box points corresponding to a second target object in a target image, where the set of target box points includes multiple initial labeled points of the second target object; in response to a first line segment formed by a first initial labeled point and a second initial labeled point in the set of target box points intersecting the first target box, determining an intersection point of the first line segment and the first target box; determining, among the first initial labeled point and the second initial labeled point, a target initial labeled point that is inside the first target box; replacing the target initial labeled point in the set of target points with the intersection point; and determining a second target box of the second target object according to the set of target box points.

[0005] In the second aspect of the embodiments of this application, a labeling device is provided. The labeling device includes a processor, a memory, and a communication circuit. The processor is respectively coupled to the memory and the communication circuit. Program data is stored in the memory, and the processor implements the steps in the above method by executing the program data in the memory.

[0006] In a third aspect of the embodiments of the present application, a labeling device is provided. The labeling device includes: an acquisition module, configured to acquire a first target box of a first target object and a set of target box point positions corresponding to a second target object in a target image, where the set of target box point positions includes a plurality of initial labeling points of the second target object, and the plurality of initial labeling points are arranged in a predetermined order; a first determination module, connected to the acquisition module, configured to determine an intersection point of a first line segment and the first target box in response to the first line segment formed by a first initial labeling point and a second initial labeling point in the set of target box point positions intersecting the first target box, where the first initial labeling point and the second initial labeling point are adjacent; a second determination module, connected to the first determination module, configured to determine a target initial labeling point located within the first target box among the first initial labeling point and the second initial labeling point; a replacement module, connected to the second determination module, configured to replace the target initial labeling point in the set of target point positions with the intersection point; a third determination module, connected to the replacement module, configured to determine a second target box of the second target object according to the set of target box point positions.

[0007] In a fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, and the computer program can be executed by a processor to implement the steps in the above method.

[0008] The beneficial effects are as follows: Different from the prior art, the present application replaces the target initial labeling points in the set of target box point positions with the corresponding intersection points, which can reduce the points in the set of target box point positions that are located within the first target box, and further reduce the intersection of the finally obtained second target box and the first target box. For the labeling personnel, there is no need for them to manually and accurately determine the points in the set of target box point positions, thus reducing the workload of the labeling personnel and effectively reducing the labeling difficulty of the labeling personnel. Description of the Drawings

[0009] To more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings, where:

[0010] Figure 1 is a flowchart of an implementation manner of the labeling method of the present application;

[0011] Figure 2 is a flowchart of determining whether the first line segment intersects the first target box in the present application;

[0012] Figure 3 is a structural diagram of the intersection of line segments;

[0013] Figure 4 Schematic flow chart of step S122 of this application;

[0014] Figure 5 is the schematic flow chart of step S120 of this application;

[0015] Figure 6 is the schematic flow chart of step S130 of this application;

[0016] Figure 7 is the schematic structural diagram of one embodiment of the target point of this application;

[0017] Figure 8 is the schematic flow chart of step S150 of this application;

[0018] Figure 9 is the schematic structural diagram of one embodiment of the marking device of this application;

[0019] Figure 10 is the schematic structural diagram of another embodiment of the marking device of this application;

[0020] Figure 11 is the schematic structural diagram of one embodiment of the computer-readable storage medium of this application. Specific embodiments

[0021] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of this application.

[0022] It should be noted that the terms "first" and "second" in this application are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of the features. In the description of this application, "a plurality" means at least two, such as two, three, etc., unless otherwise specifically defined. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or devices.

[0023] Refer to Figure 1 , Figure 1FIG. 0 is a schematic flowchart of an implementation of the annotation method of the present application. This method is executed by an annotation device, which can be any device with algorithm processing capabilities such as a mobile phone or a computer. This method includes:

[0024] S110: Obtain the first target box of the first target object in the target image and the set of target box point positions corresponding to the second target object.

[0025] Specifically, the first target object and the second target object can be objects of the same type. For example, they are both people or both vehicles, etc. They can also be objects of different types. For example, the first target is a person and the second target object is a vehicle, etc.

[0026] Among them, the first target box of the first target object has been pre-annotated and is determined, and there will be no subsequent changes to it.

[0027] At the same time, the set of target box point positions corresponding to the second target object includes multiple initial annotation points of the second target object. Among them, the multiple initial annotation points included in the set of target box point positions can be recognized by the annotation device. At this time, the annotation device recognizes the second target object to obtain multiple initial annotation points. Or, the multiple initial annotation points can also be input by the user. For example, the user clicks multiple points along the contour of the second target object on the target image, and the annotation device recognizes the points clicked by the user to obtain multiple initial annotation points.

[0028] Among them, the initial target box of the second target object can be determined according to the multiple initial annotation points in the set of target box point positions, but the accuracy of this initial target box is low. Especially when the second target object is blocked by the first target object, this initial target box may intersect with the first target box of the first target object. Therefore, in order to avoid this phenomenon, steps S120 - S150 are executed.

[0029] Among them, in the set of target box point positions, the multiple initial annotation points are arranged in a predetermined order. Specifically, at this time, after connecting all adjacent two initial annotation points in sequence, the initial target box of the second target object can be obtained.

[0030] S120: In response to the first line segment formed by the first initial annotation point and the second initial annotation point in the set of target box point positions intersecting the first target box, determine the intersection point of the first line segment and the first target box.

[0031] Specifically, the first initial annotation point and the second initial annotation point are adjacent, that is, they are adjacent point positions, and the first line segment formed by them is a part of the initial target box of the second target object. At this time, for each adjacent two initial annotation points in the set of target box point positions, steps S120 - step S140 can be executed.

[0032] For the sake of convenience of explanation, the following will be described by taking steps S120 - S140 for the first initial annotation point and the second initial annotation point.

[0033] Among them, for the first line segment formed by the first initial annotation point and the second initial annotation point, it is judged whether it intersects with the first target box. If it intersects, the intersection point of the two is determined. The process of judging whether the first line segment intersects with the first target box and the process of determining the intersection point of the two can be seen in the following introduction.

[0034] S130: Among the first initial annotation point and the second initial annotation point, determine the target initial annotation point that is within the first target box.

[0035] Specifically, since the first line segment intersects with the first target box, among the two points of the first initial annotation point and the second initial annotation point, there must be a point inside the first target box. At this time, the point within the first target box is defined as the target initial annotation point. The specific process of determining the target initial annotation point can be seen in the following introduction.

[0036] S140: Replace the target initial annotation point in the target point set with the intersection point.

[0037] Specifically, because there must be a target initial point within the first target box among the two points of the first initial annotation point and the second initial annotation point, if the first line segment continues to be used as a part of the initial target box of the second target object, then this initial target box will definitely intersect with the first target box. Therefore, to avoid this phenomenon, the intersection point of the first line segment and the first target box is used to replace the target initial annotation point inside the first target box.

[0038] It should be noted that after the replacement, the position order of the intersection point in the target box point set is the same as the position order of the target initial annotation point in the target box point set. For example, if the target initial annotation point is the 5th point in the target box point set, then after the replacement, the intersection point is still the 5th point in the target box point set.

[0039] S150: Determine the second target box of the second target object according to the target box point set.

[0040] Specifically, after steps S120 to S140, at least some of the initial annotation points in the target box point set that are located within the first target box can be replaced with the corresponding intersection points, so as to reduce the intersection of the second target box and the first target box. Thus, it is not necessary for the annotator to accurately annotate the second target box of the second target object, which can reduce the workload of the annotator and effectively reduce the annotation difficulty of the annotator.

[0041] It should be noted that for a certain initial annotation point in the set of target box points, if the initial annotation point is replaced with the corresponding intersection point, it is not necessary to determine whether the line segment formed by the intersection point and other initial annotation points intersects with the first target box.

[0042] For the sake of easy understanding, an example is given for illustration. Assume that the set of target box points includes point A, point B, point C, and point D. When point B is replaced with the corresponding intersection point, directly determine whether the line segment formed by point C and point D intersects with the first target box.

[0043] In this embodiment, refer to Figure 2 , the process of determining whether the first line segment intersects with the first target box specifically includes:

[0044] S121: Sequentially determine whether the first line segment intersects with the second line segments that make up the first target box.

[0045] Specifically, the first target box can be split into multiple line segments, and this line segment is defined as the second line segment. Among all the second line segments, sequentially determine whether the second line segment intersects with the first line segment.

[0046] S122: If it is determined that there is a second line segment that intersects with the first line segment, determine that the first line segment intersects with the first target box.

[0047] Specifically, if it is determined that there is a second line segment that intersects with the first line segment, determine that the first line segment intersects with the first target box; if all the second line segments do not intersect with the first line segment, determine that the first line segment does not intersect with the first target box.

[0048] In other embodiments, the process of determining whether the first line segment intersects with the first target box can also be: determine whether the first initial annotation point and the second initial annotation point are within the first target box. If one of the first initial annotation point and the second initial annotation point is within the first target box and the other is not within the first target box, determine that the first line segment intersects with the first target box; otherwise, determine that the first line segment does not intersect with the first target box.

[0049] Among them, for a point, the specific process of determining whether it is within the first target box can be seen in the following introduction.

[0050] In this embodiment, the second line segment that intersects with the first line segment is defined as the target second line segment; when replacing the target initial annotation point in the set of target points with the intersection point, the method further includes: among the two endpoints of the target second line segment, determine the target endpoint that is within the third target box, and add the target endpoint to the set of target box points. Among them, the third target box is determined according to the points in the set of target box points.

[0051] Specifically, for ease of understanding, it is described with examples:

[0052] In Figure 3 it is the case that line segment AB and line segment CD intersect, and the intersection point is E. The two endpoints of line segment AB are respectively the first initial annotation point and the second initial annotation point of the second target box, that is, line segment AB is the first line segment, and line segment CD is the target second line segment. At the same time, it is assumed that point A is the target initial annotation point. Then, while replacing point A with point E in the target box point set, it is also necessary to determine which of point C and point D is in the third target box. If point C is in the third target box, then point C is added to the target box point set. If point D is in the third target box, then point D is added to the target box point set.

[0053] Among them, the third target box is determined according to the points in the target box point set. It can be understood that since the target box point set is constantly updated, the third target box is the latest target box of the second target object.

[0054] Among them, if point D is the target endpoint, then both E and point D exist in the final target box point set, and line segment ED is one side of the second target box. Since line segment DE is a part of line segment CD, line segment DE is the common side of the first target box and the second target box.

[0055] That is to say, the purpose of searching for the target endpoint among the two endpoints of the target second line segment is to determine the common side of the first target box and the second target box, so as to further reduce the intersection of the first target box and the second target box.

[0056] It should be noted that in other embodiments, it is also possible to only replace the target initial annotation point with the intersection point, without searching for the target endpoint among the two endpoints of the target second line segment, nor adding the target endpoint to the target box point set.

[0057] Referring to Figure 4 , the specific process of step S122 for determining whether the first line segment intersects the second line segment includes:

[0058] S1221: Determine the normal line of the second line segment.

[0059] S1222: Respectively determine the projection distances from the two endpoints of the first line segment and the endpoint of the second line segment closest to the normal line to the normal line.

[0060] S1223: Determine whether the first line segment intersects the second line segment according to the projection distances.

[0061] Specifically, for ease of understanding, in combination with Figure 3, taking line segment AB as the first line segment and line segment CD as the second line segment, and assuming that the coordinates of points A, B, C, and D are (ax, ay), (bx, by), (cx, cy), and (dx, dy) respectively, the process of determining whether line segment AB intersects line segment CD includes:

[0062] First, calculate the normal coordinates of line segment CD as (nx, ny), where nx = dy - cy and ny = cx - dx;

[0063] Then, determine the projection distances of points A, B, and C on the normal of line segment CD as: distC = cx * nx + cy * ny, disA = ax * nx + ay * ny, distB = bx * nx + by * ny, where distC is the projection distance of point C on the normal of line segment CD, disA is the projection distance of point A on the normal of line segment CD, and distB is the projection distance of point B on the normal of line segment CD;

[0064] Finally, if the calculation result of (distA - distC) * (distB - distC) is less than 0, it is determined that line segment AB intersects line segment CD; otherwise, it is determined that line segment AB does not intersect line segment CD.

[0065] Refer to Figure 5 , in this embodiment, the process of step S120 for determining the intersection point of the first line segment and the first target box includes:

[0066] S1201: Determine the normals of the first line segment and the target second line segment respectively.

[0067] S1202: Determine the intersection point of the first line segment and the target second line segment according to the first line segment, the target second line segment, the normal of the first line segment, and the normal of the target second line segment, and determine the intersection point of the first line segment and the first target box as the intersection point of the first line segment and the target second line segment.

[0068] Continue to combine Figure 3 , still taking line segment AB as the first line segment and line segment CD as the second line segment, referring to the above process of determining the normal coordinates of line segment CD, the normal coordinates of line segment AB can be determined, denoted as (mx, my). At the same time, after obtaining the projection distances of points C, A, and B as distC, disA, and distB respectively according to the above method, the coordinates (x, y) of the intersection point E are calculated according to the following formula:

[0069] num = (distA - distC) / (mx * ny - my * nx); x = ax + num * my; y = ay - num * mx.

[0070] Refer to Figure 6, step S130 specifically includes:

[0071] S131: Determine the included angles of each second line segment that forms the first target box corresponding to the first initial annotation point. Among them, the included angle of the second line segment corresponding to the first initial annotation point is the included angle between the connection lines of the first initial annotation point and the two endpoints of the second line segment.

[0072] S132: In response to the sum of the included angles being equal to 360°, determine the first initial annotation point as the target initial annotation point.

[0073] S133: In response to the sum of the included angles not being equal to 360°, determine the second initial annotation point as the target initial annotation point.

[0074] Specifically, for easy understanding, combined with Figure 7 Illustrate steps S131 - S133:

[0075] Quadrilateral abcd is the first target box. Line segments ab, bc, cd, and da are all second line segments. When it is necessary to determine whether point o is within the first target box, respectively determine ∠1, ∠2, ∠3, and ∠4 of point o corresponding to line segments ab, bc, cd, and da. If the sum of the degrees of ∠1, ∠2, ∠3, and ∠4 is equal to 360°, then point o is within the first target box; otherwise, determine that point o is not within the first target box.

[0076] Among them, the process of calculating ∠1 is as follows:

[0077] ∠1 = degrees(acos(ab * ab - bo * bo - ao * ao) / (-2 * ao * bo));

[0078] Among them, ab, bc, cd, and da are the lengths of line segments ab, bc, cd, and da, and bo, ao are the lengths of line segments bo, ao.

[0079] Among them, the calculation processes of ∠2, ∠3, and ∠4 can refer to that of ∠1, which will not be elaborated here.

[0080] Refer to Figure 8 , in this embodiment, step S150 specifically includes:

[0081] S151: Remove the points in the target box point set that are within the first target box.

[0082] S152: Use the remaining points in the target box point set to draw the second target box.

[0083] Specifically, combined with Figure 8, although after steps S120 - S140, the target initial annotation points in the target box point set have been replaced with the corresponding intersection points, in order to avoid that there are still points in the target box point set that are within the first target box, for each point in the target box point set, it is determined whether it is within the first target box. If it is, it is removed to further refine the target box point set.

[0084] The specific process of determining whether a point is within the first target box has been introduced above and will not be elaborated here.

[0085] It should be noted that in other embodiments, step S150 can also directly draw the second target box using the points in the target box point set without removing the points within the first target box.

[0086] In an application scenario, before further drawing the second target box in step S150, it is also possible to further remove the points in the target box point set that are within other annotation boxes to ensure that there are no points within any other annotation box in the final target box point set.

[0087] Refer to Figure 9 , Figure 9 is a schematic structural diagram of an embodiment of the annotation device of the present application. The annotation device 200 includes an acquisition module 210, a first determination module 220, a second determination module 230, a replacement module 240, and a third determination module 250 that are connected in sequence.

[0088] The acquisition module 210 is used to acquire the first target box of the first target object and the target box point set corresponding to the second target object in the target image. Among them, the target box point set includes multiple initial annotation points of the second target object, and the multiple initial annotation points are arranged in a predetermined order.

[0089] The first determination module 220, connected to the acquisition module 210, is used to determine the intersection points of the first line segment formed by the first initial annotation point and the second initial annotation point in the target box point set with the first target box, where the first initial annotation point and the second initial annotation point are adjacent.

[0090] The second determination module 230, connected to the first determination module 220, is used to determine the target initial annotation points within the first target box among the first initial annotation point and the second initial annotation point.

[0091] The replacement module 240, connected to the second determination module 230, is used to replace the target initial annotation points in the target point set with the intersection points.

[0092] The third determination module 250, connected to the replacement module 240, is configured to determine a second target box of the second target object according to the set of target box positions.

[0093] Wherein, when the annotation device 200 is working, it can execute the specific steps in any of the above embodiments. For the detailed steps, reference can be made to the relevant content above, which will not be elaborated here.

[0094] Wherein, the annotation device 200 can be any device with algorithm processing capabilities such as a mobile phone or a computer, which will not be elaborated here.

[0095] Refer to Figure 10 , Figure 10 FIG. is a schematic structural diagram of another embodiment of the annotation device of the present application. The annotation device 300 includes a processor 310, a memory 320, and a communication circuit 330. The processor 310 is respectively coupled to the memory 320 and the communication circuit 330. Program data is stored in the memory 320. The processor 310 realizes the steps in the method of any of the above embodiments by executing the program data in the memory 320. For the detailed steps, reference can be made to the above embodiments, which will not be elaborated here.

[0096] Wherein, the annotation device 300 can be any device with image processing capabilities such as a computer or a mobile phone, and is not limited here.

[0097] Wherein, the annotation device 300 can be any device with algorithm processing capabilities such as a mobile phone or a computer, which will not be elaborated here.

[0098] Refer to Figure 11 , Figure 11 FIG. is a schematic structural diagram of an embodiment of a computer-readable storage medium of the present application. The computer-readable storage medium 400 stores a computer program 410, and the computer program 410 can be executed by a processor to realize the steps in any of the above methods.

[0099] Wherein, the computer-readable storage medium 400 can specifically be a device such as a USB flash drive, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk, or an optical disc that can store the computer program 410, or it can also be a server storing the computer program 410. The server can send the stored computer program 410 to other devices for running, or it can also run the stored computer program 410 by itself.

[0100] The above are only embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present application, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present application.

Claims

1. A marking method, characterized in that, The method includes: Obtaining a first target box of a first target object and a set of target box points corresponding to a second target object in a target image, where the set of target box points includes a plurality of initial annotation points of the second target object, and the plurality of initial annotation points are arranged in a predetermined order; In response to a first line segment formed by a first initial annotation point and a second initial annotation point in the set of target box points intersecting the first target box, determining an intersection point of the first line segment and the first target box, where the first initial annotation point and the second initial annotation point are adjacent; Determining a target initial annotation point that is within the first target box among the first initial annotation point and the second initial annotation point; Replacing the target initial annotation point in the set of target box points with the intersection point; Determining a second target box of the second target object according to the set of target box points.

2. The method according to claim 1, wherein Before the step of, in response to a first line segment formed by a first initial annotation point and a second initial annotation point in the set of target box points intersecting the first target box, determining an intersection point of the first line segment and the first target box, further including: Sequentially determining, for a second line segment forming the first target box, whether the first line segment intersects the second line segment; If it is determined that there is a second line segment that intersects the first line segment, determining that the first line segment intersects the first target box.

3. The method according to claim 2, wherein Defining the second line segment that intersects the first line segment as a target second line segment; While replacing the target initial annotation point in the set of target box points with the intersection point, further including: Determining a target end point that is within a third target box among two end points of the target second line segment, where the third target box is determined according to points in the set of target box points; Adding the target end point to the set of target box points.

4. The method according to claim 2, wherein The step of determining whether the first line segment intersects the second line segment includes: Determining a normal line of the second line segment; Respectively determining projection distances from two end points of the first line segment and an end point of the second line segment closest to the normal line to the normal line; Determining whether the first line segment intersects the second line segment according to the projection distances.

5. The method according to claim 2, wherein Defining the second line segment that intersects the first line segment as a target second line segment; The step of determining the intersection point of the first line segment and the first target box includes: Respectively determining normal lines of the first line segment and the target second line segment; Determining an intersection point of the first line segment and the target second line segment according to the first line segment, the target second line segment, the normal line of the first line segment, and the normal line of the target second line segment, and determining the intersection point of the first line segment and the target second line segment as the intersection point of the first line segment and the first target box.

6. The method according to claim 1, wherein The step of determining a target initial annotation point that is within the first target box among the first initial annotation point and the second initial annotation point includes: Determine the included angle of each second line segment that forms the first target box corresponding to the first initial annotation point, where the included angle corresponding to the second line segment by the first initial annotation point is the included angle between the lines connecting the first initial annotation point and the two endpoints of the second line segment; In response to the sum of the included angles being equal to 360°, determine the first initial annotation point as the target initial annotation point; In response to the sum of the included angles not being equal to 360°, determine the second initial annotation point as the target initial annotation point.

7. The method according to claim 1, characterized in that, The step of determining the second target box of the second target object according to the target box point set further includes: Remove the points in the target box point set that are within the first target box; Use the remaining points in the target box point set to draw the second target box.

8. A labeling device, characterized in that, The annotation device includes a processor, a memory, and a communication circuit. The processor is respectively coupled to the memory and the communication circuit. Program data is stored in the memory, and the processor implements the steps in the method according to any one of claims 1-7 by executing the program data in the memory.

9. A marking device, characterized in that, The annotation device includes: An acquisition module, configured to acquire the first target box of the first target object and the target box point set corresponding to the second target object in the target image, where the target box point set includes multiple initial annotation points of the second target object, and the multiple initial annotation points are arranged in a predetermined order; A first determination module, connected to the acquisition module, configured to determine the intersection point of the first line segment formed by the first initial annotation point and the second initial annotation point in the target box point set and the first target box in response to the first line segment intersecting the first target box, where the first initial annotation point and the second initial annotation point are adjacent; A second determination module, connected to the first determination module, configured to determine the target initial annotation point that is within the first target box among the first initial annotation point and the second initial annotation point; A replacement module, connected to the second determination module, configured to replace the target initial annotation point in the target box point set with the intersection point; A third determination module, connected to the replacement module, configured to determine the second target box of the second target object according to the target box point set.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and the computer program can be executed by a processor to implement the steps in the method according to any one of claims 1-7.

Images