Visual model training data set expansion method, device, equipment and storage medium

Abstract

The application provides a visual model training data set expansion method and device, equipment and storage medium, including: generating a specified motion path in a scene corresponding to a target video stream, and setting attribute information corresponding to the specified motion path; obtaining shooting information of each point on the specified motion path, the shooting information including a shooting angle and a shooting distance; based on an image set in the target video stream that needs to add a specified motion object, based on a shooting information set and a 3D model corresponding to the specified motion object, obtaining a two-dimensional image corresponding to the specified motion object at each point, and placing the obtained two-dimensional image at a corresponding position of a corresponding image in the target video stream. The application can add a motion trajectory of a moving object in the target video stream.

Description

Technical Field

[0001] This invention relates to the field of image processing, and in particular to a method, apparatus, and storage medium for expanding a visual model training dataset. Background Technology

[0002] In some application scenarios, such as the training of a prediction model that needs to predict the motion path of a moving object, a large amount of sample data is required to train the model. However, due to certain reasons, not enough sample data can be obtained. Therefore, in order to ensure that the model can be trained effectively, the existing sample data needs to be expanded. Summary of the Invention

[0003] To address the aforementioned technical problems, the technical solution adopted by this invention is as follows:

[0004] According to a first aspect of the present invention, a method for expanding a visual model training dataset is provided, the method comprising the following steps:

[0005] S100: Obtain the video stream to which the specified motion object needs to be added, and use it as the target video stream.

[0006] S200, a specified motion path is generated in the scene corresponding to the target video stream, and attribute information corresponding to the specified motion path is set, the attribute information including motion time and motion form.

[0007] S300, acquire the shooting information of each point on the specified motion path. The shooting information includes the shooting angle and shooting distance, wherein the shooting angle of a point is the angle between the line connecting the point to the corresponding shooting device and the shooting line of the shooting device, and the shooting distance of a point is the distance between the point and the corresponding shooting device; obtain the shooting information set C={C1, C2, ..., C...} corresponding to the motion path. i , ..., C n}, the i-th captured information C in C i For the image capture information of the i-th point on the specified motion path, C i =(α i d i ), where i ranges from 1 to n, n is the number of points on the specified motion path, and α i Let d be the shooting angle corresponding to the i-th point. i Let be the shooting distance corresponding to the i-th point.

[0008] S400: Based on the image set of the target video stream to which a specified moving object needs to be added, based on C and the 3D model corresponding to the specified moving object, obtain the two-dimensional image of the specified moving object at each point, and place the obtained two-dimensional image at the corresponding position of the corresponding image in the target video stream.

[0009] According to a second aspect of the present invention, a visual model training dataset expansion device is provided, the device comprising:

[0010] The first acquisition module is used to acquire the video stream to which the specified motion object needs to be added, and use it as the target video stream.

[0011] The path generation module is used to generate a specified motion path in the scene corresponding to the target video stream, and set the attribute information corresponding to the specified motion path, including motion time and motion form.

[0012] The second acquisition module is used to acquire shooting information for each point on the specified motion path. The shooting information includes shooting angle and shooting distance. The shooting angle of a point is the angle between the line connecting the point to the corresponding shooting device and the shooting line of sight of the shooting device. The shooting distance of a point is the distance between the point and the corresponding shooting device. The module obtains a shooting information set C = {C1, C2, ..., C...} corresponding to the motion path. i , ..., C n}, the i-th captured information C in C i For the image capture information of the i-th point on the specified motion path, C i =(α i d i ), where i ranges from 1 to n, n is the number of points on the specified motion path, and α i Let d be the shooting angle corresponding to the i-th point. i Let be the shooting distance corresponding to the i-th point.

[0013] The addition module is used to obtain a two-dimensional image of the specified moving object at each point based on the image set of the target video stream to which a specified moving object needs to be added, based on C and the 3D model corresponding to the specified moving object, and to place the obtained two-dimensional image into the corresponding position of the corresponding image in the target video stream.

[0014] According to a third aspect of the present invention, an electronic device is provided, including a processor and a memory; the processor executes the steps of the method described in the first aspect of the present invention by invoking a program or instructions stored in the memory.

[0015] According to a fourth aspect of the present application, there is provided a non-transitory computer readable storage medium storing a program or instructions for causing a computer to perform the steps of the method according to the first aspect of the present application.

[0016] The present application has at least the following beneficial effects:

[0017] The visual model training data set expansion method provided by the embodiments of the present application can add the motion trajectory of the moving object in the training data set, and can expand the sample data.

[0018] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor on the basis of these drawings.

[0020] Figure 1 The flow chart of the visual model training data set expansion method provided by the embodiments of the present application. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described clearly and completely in the following description with reference to the drawings of the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terminology used in the description of the present application herein only for the purpose of describing the specific embodiments, and is not intended to limit the present application. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0023] It is to be understood that some of the example embodiments are described in terms of a process or method depicted as a flowchart. Although each step in a flowchart can be identified with a reference number and can describe an operation, action, or decision, the reference numbers can correspond, in some embodiments, to portions of code. Although the flowcharts depict a particular order of operations, the order of the operations can be changed so that

[0024] The embodiment of the present application provides a visual model training data set expansion method, which comprises the following steps as shown in the figure: Figure 1 The embodiment of the present application provides a visual model training data set expansion method, which comprises the following steps as shown in the figure:

[0025] S100, acquiring a video stream currently needing to add a specified moving object as a target video stream.

[0026] In the embodiment of the present application, the target video stream can be a video stream specified by a user, and can be a video stream obtained by continuously shooting a corresponding scene. For example, a video stream obtained by continuously shooting a traffic intersection or a video stream obtained by continuously shooting a certain road.

[0027] In the embodiment of the present application, the specified moving object can be a moving object, for example, a pedestrian, a vehicle, an airplane, etc.

[0028] S200, generating a specified moving path in a scene corresponding to the target video stream, and setting attribute information corresponding to the specified moving path, wherein the attribute information comprises a moving time and a moving form.

[0029] In the embodiment of the present application, the specified moving path can be an arbitrary path specified by a user. In a specific implementation, the specified moving path can be specified in a first frame image of the target video stream.

[0030] In the embodiment of the present application, the moving time refers to a time required from a starting point position to an ending point position of the specified moving path, and can be a time specified by a user. The moving form refers to a moving speed type from the starting point position to the ending point position of the specified moving path, and includes uniform motion, accelerated motion, etc. In an illustrative embodiment of the present application, the moving form can be uniform motion.

[0031] S300, acquiring shooting information of each point on the specified moving path, wherein the shooting information comprises a shooting angle and a shooting distance, the shooting angle of a certain point refers to an included angle between a line connecting the point and a corresponding shooting device and a shooting sight line of the shooting device, and the shooting distance of the certain point refers to a relative distance between the point and the corresponding shooting device; obtaining a shooting information set C={C1, C2, …, C i , …, C n} corresponding to the moving path, wherein the i th shooting information Ci C is the shooting information of the i-th point on the specified motion path i = (a i , d i ), i is an integer from 1 to n, n is the number of points on the specified motion path, a i is the shooting angle corresponding to the i-th point, and d i is the shooting distance corresponding to the i-th point.

[0032] Since the position of the camera and the shooting line of sight are relatively fixed during shooting, the shooting angle and the shooting distance of each pixel point in each frame of image in the video stream are basically unchanged. Therefore, the shooting angle and the shooting distance of each pixel point in the image can be obtained according to existing knowledge.

[0033] S400, based on the image set in the target video stream in which the specified motion object needs to be added, based on C and the 3D model corresponding to the specified motion object, obtaining the two-dimensional image corresponding to each point of the specified motion object, and placing the obtained two-dimensional image in the corresponding position of the corresponding image in the target video stream.

[0034] Further, in an illustrative embodiment, S400 can specifically include:

[0035] S410, setting a variable counter j = 1.

[0036] S420, if j≤m, executing S430, otherwise executing S480; m is the frame number of the images in the image set;

[0037] S430, for the j-th frame image IMG j in the image set, obtaining the time interval△t j between IMG j-1 and the previous frame image IMG j（j-1） .

[0038] In the embodiment of the application, each frame of image corresponds to a shooting time, and△t j（j-1） is the difference between the shooting times of IMG j and IMG j-1 .

[0039] S440, based on the start position, the end position, the motion time, the motion form corresponding to the specified motion path, the position G j-1 of the added point corresponding to IMG j-1 , and△t j（j-1） , obtaining the position G j of the added point in IMG j in which the specified motion object needs to be added; G1 is the start position corresponding to the specified motion path.

[0040] Those skilled in the art will understand that, assuming a specified moving object moves along a specified motion path, based on the starting position, ending position, motion time, motion form, and IMG corresponding to the specified motion path... j-1 The corresponding location of the added point G j-1 and △t j（j-1） This allows you to obtain the position of a specified moving object from its previous position G. j-1 After time △t j（j-1） The position of G1 in the specified motion path is then determined, and any method for obtaining G1 falls within the protection scope of this invention.

[0041] S450, based on G j Obtain the corresponding shooting information from C, and based on G j The corresponding shooting information controls the virtual camera corresponding to the 3D model to capture images of specified moving objects in the 3D model, obtaining the corresponding two-dimensional images as G. j The corresponding two-dimensional image DIMG j .

[0042] In this embodiment of the invention, the virtual camera and the shooting device corresponding to the target video stream have the same parameters.

[0043] Those skilled in the art will know that any G-based j The methods for controlling the virtual camera corresponding to the 3D model to capture a specified moving object in the 3D model and obtain a corresponding two-dimensional image, based on the corresponding shooting information, are all within the protection scope of this invention.

[0044] In this embodiment of the invention, since it is based on G j The corresponding shooting information controls the virtual camera corresponding to the 3D model to shoot the specified moving object in the 3D model and obtain the corresponding two-dimensional image, so that the moving object added to the target video stream can be relatively realistic.

[0045] S460, from DIMG j The specified motion object is segmented from the image and added to the IMG. j G j Place.

[0046] In this embodiment of the invention, existing image overlay methods can be used to add the segmented, specified moving object to the IMG. j G j Place.

[0047] S470 will add the specified motion image to the IMG. j Add to the current intermediate image set; set j=j+1 and execute S420; the initial value of the current intermediate image set is empty.

[0048] S480, taking the current intermediate image set as a target image set corresponding to the target video stream.

[0049] In another illustrative embodiment of the present application, S400 can specifically include:

[0050] S411, generating a camera movement path corresponding to a virtual camera corresponding to the 3D model based on C.

[0051] As known by those skilled in the art, any method of generating a camera movement path corresponding to a virtual camera corresponding to the 3D model based on C falls within the protection scope of the present application.

[0052] S412, controlling the virtual camera to take pictures based on the camera movement path, to obtain a corresponding image set IMGS={DIMG1, DIMG2, …, DIMG i , …, DIMG m}, DIMG i is a two-dimensional image taken by the virtual camera based on C. i

[0053] S413, setting a variable counter j=1.

[0054] S414, if j≤m, executing S415, otherwise, executing S420; m is the frame number of images in the image set.

[0055] S415, for the jth image IMG j in the image set, obtaining a time interval△t j between IMG j-1 and the previous image IMG j（j-1） .

[0056] S416, based on the start position, the end position, the motion time, the motion form, the position G j-1 of the added point corresponding to IMG j-1 and△t j（j-1） , obtaining the position G j of the added point in IMG j which needs to add the specified motion object; G1 is the start position corresponding to the specified motion path.

[0057] S417, obtaining a two-dimensional image corresponding to IMG j from IMGS as a two-dimensional image DIMG j corresponding to IMG j .

[0058] In the embodiment of the present application, G​j acquire the corresponding two-dimensional image from the IMGS corresponding to the G j corresponding two-dimensional image as IMG j corresponding two-dimensional image DIMG j .

[0059] S418, segment the specified moving object from the DIMG j , and add the segmented specified moving object to the G j of IMG j .

[0060] S419, add the IMG j with the specified moving image to the current intermediate image set; set j = j + 1, and execute S414; the initial value of the current intermediate image set is empty.

[0061] S420, take the current intermediate image set as the target image set corresponding to the target video stream.

[0062] The specific implementation of S413 to S416 can refer to the specific implementation of the foregoing S410 to S440, and the specific implementation of S418 to S420 can refer to the specific implementation of the foregoing S460 to S480.

[0063] Compared with the foregoing embodiments, in the embodiment, the movement path of the photographing device is first generated based on C, then the 3D model is photographed according to the movement path, the 3D model is mapped into a two-dimensional space, and a two-dimensional image set corresponding to each point on the specified movement path is obtained, so that the corresponding two-dimensional image can be directly acquired from the image set when needed, and efficiency can be improved.

[0064] Further, in the embodiment of the application, the center of the bounding box of the specified moving object is the corresponding added point.

[0065] Further, in the embodiment of the application, after S460, the following steps are further included:

[0066] S462, identify the target object in IMG j , and obtain the corresponding recognition frame; the target object includes the specified moving object.

[0067] In the embodiment of the application, the target object in the image can be identified by using a target detection algorithm such as a YOLO series target detection algorithm. The specific type of the target object can be set based on actual needs.

[0068] S464, if there is an overlapping area between the recognition frame corresponding to the segmented specified moving object in DIMG j and any recognition frame r in IMG j , if dj <d r , it is indicated that the specified moving object is located in front of the target object corresponding to the recognition box relative to the photographing device, and the region belonging to the recognition box r in the overlapping region needs to be occluded, if d j >d r , it is indicated that the specified moving object is located in the rear of the target object corresponding to the recognition box relative to the photographing device, and the region belonging to the recognition box corresponding to the specified moving object segmented out in DIMG j needs to be occluded, d r is the distance between the recognition box r and the corresponding photographing device, and j is valued from 1 to f(i)-1, and f(i) is the number of recognition boxes in IMG j ; S470 is performed.

[0069] In the embodiments of the present application, any method of occluding the region in the image belongs to the protection scope of the present application.

[0070] Further, in the embodiments of the present application, after S418, the following steps are further included:

[0071] S4181, the target object in IMG j is recognized to obtain the corresponding recognition box; the target object includes the specified moving object.

[0072] S4182, if there is an overlapping region between the recognition box corresponding to the specified moving object segmented out in DIMG j and any recognition box r in IMG j , if d j <d r , the region belonging to the recognition box r in the overlapping region is occluded, if d j >d r , the region belonging to the recognition box corresponding to the specified moving object segmented out in DIMG j in the overlapping region is occluded, d r is the distance between the recognition box r and the corresponding photographing device, and j is valued from 1 to f(i)-1, and f(i) is the number of recognition boxes in IMG j ; S419 is performed.

[0073] The visual model training data set expansion method provided by the embodiments of the present application can generate a video stream added with the motion trajectory of the moving object, and can expand the sample data used for model prediction, because the 3D model of the specified moving object is photographed based on the relative distance and the photographing angle of each point of the specified motion path of the scene corresponding to the target video stream relative to the photographing device, and the corresponding two-dimensional image is obtained, and then added to the corresponding image in the target video stream.

[0074] Based on the same inventive concept, the embodiment of the present application provides a visual model training data set expansion device, which comprises:

[0075] The first acquisition module is configured to acquire a video stream in which a specified moving object needs to be added as a target video stream.

[0076] The path generation module is configured to generate a specified moving path in a scene corresponding to the target video stream and set attribute information corresponding to the specified moving path, wherein the attribute information comprises a moving time and a moving form.

[0077] The second acquisition module is configured to acquire shooting information of each point on the specified moving path, wherein the shooting information comprises a shooting angle and a shooting distance between a corresponding shooting device; and obtain a shooting information set C={C1, C2, …, C i , …, C n} corresponding to the moving path, wherein the i-th shooting information C i in the set C is shooting information of the i-th point on the specified moving path, C i =(α i , d i ), i is an integer from 1 to n, n is the number of points on the specified moving path, α i is the shooting angle corresponding to the i-th point, and d i is the shooting distance corresponding to the i-th point.

[0078] The adding module is configured to acquire a two-dimensional image of the specified moving object at each point based on the set C and a 3D model corresponding to the specified moving object based on an image set in which the specified moving object needs to be added in the target video stream, and place the acquired two-dimensional image in a corresponding position of a corresponding image in the target video stream.

[0079] The device can be used to execute the method shown in the embodiment shown in Figure 1 , therefore, the functions and the like that can be achieved by the functional modules of the device can refer to the description of the embodiment shown in Figure 1 , and no more description is made.

[0080] The embodiment of the present application also provides an electronic device, which comprises at least one processor and a memory connected with the at least one processor in communication; wherein the memory stores instructions executable by the at least one processor, and the instructions are configured to execute the method described in the embodiment of the present application.

[0081] The embodiment of the present application also provides a non-transitory computer readable storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the method described in the embodiment of the present application.

[0082] It should be understood that the various forms of flow shown above can be used to reorder, add, or delete steps. For example, the steps described in the present application can be performed in parallel, in series, or in a different order, as long as the desired results of the technology disclosed in the present application can be achieved, which is not limited herein.

[0083] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for expanding a visual model training dataset, characterized in that, The method includes the following steps: S100: Obtain the video stream to which the specified motion object needs to be added, and use it as the target video stream; S200, generate a specified motion path in the scene corresponding to the target video stream, and set the attribute information corresponding to the specified motion path, the attribute information including motion time and motion form; S300, acquire the shooting information of each point on the specified motion path. The shooting information includes the shooting angle and shooting distance, wherein the shooting angle of a point is the angle between the line connecting the point to the corresponding shooting device and the shooting line of the shooting device, and the shooting distance of a point is the distance between the point and the corresponding shooting device; obtain the shooting information set C={C1, C2, ..., C...} corresponding to the motion path. i , ..., C n }, the i-th captured information C in C i For the image capture information of the i-th point on the specified motion path, C i =(α i d i ), where i ranges from 1 to n, n is the number of points on the specified motion path, and α i Let d be the shooting angle corresponding to the i-th point. i The shooting distance corresponding to the i-th point; S400: Based on the image set of the target video stream to which a specified moving object needs to be added, based on C and the 3D model corresponding to the specified moving object, obtain the two-dimensional image of the specified moving object at each point, and place the obtained two-dimensional image at the corresponding position of the corresponding image in the target video stream.

2. The method according to claim 1, characterized in that, The S400 specifically includes: S410, set variable counter j=1; S420, if j≤m, execute S430, otherwise execute S480; m is the number of frames in the image set; S430, for the j-th frame image IMG in the image set j Get IMG j And the previous frame image IMG j-1 The time interval Δt between j（j-1） ; S440, based on the starting position, ending position, movement time, movement type, and IMG corresponding to the specified movement path. j-1 The corresponding location of the added point G j-1 and △t j（j-1） Get IMG j The location G of the point to be added for the specified motion object needs to be added. j G1 is the starting position corresponding to the specified motion path; S450, based on G j Obtain the corresponding shooting information from C, and based on G j The corresponding shooting information controls the virtual camera corresponding to the 3D model to capture images of specified moving objects in the 3D model, obtaining the corresponding two-dimensional images as G. j The corresponding two-dimensional image DIMG j ; S460, from DIMG j The specified motion object is segmented from the image and added to the IMG. j G j Place; S470 will add the specified motion image to the IMG. j Add to the current intermediate image set; set j=j+1 and execute S420; the initial value of the current intermediate image set is empty; S480, the current intermediate image set is used as the target image set corresponding to the target video stream.

3. The method according to claim 1, characterized in that, The S400 specifically includes: S411, Based on C, generate the camera movement path corresponding to the virtual camera of the 3D model; S412, based on the camera movement path, control the virtual camera to take pictures and obtain the corresponding image set IMGS={DIMG1, DIMG2, ..., DIMG...} i , ..., DIMG m }, DIMG i For virtual cameras based on C i Two-dimensional images obtained from photography; S413, Set variable counter j=1; S414, if j≤m, execute S415, otherwise execute S420; m is the number of frames in the image set; S415, for the j-th frame image IMG in the image set j Get IMG j And the previous frame image IMG j-1 The time interval Δt between j（j-1） ; S416, based on the starting position, ending position, movement time, movement type, and IMG corresponding to the specified movement path. j-1 The corresponding location of the added point G j-1 and △t j（j-1） Get IMG j The location G of the point to be added for the specified motion object needs to be added. j G1 is the starting position corresponding to the specified motion path; S417, Obtain from IMG j The corresponding two-dimensional image, as IMG j The corresponding two-dimensional image DIMG j ; S418, from DIMG j The specified motion object is segmented from the image and added to the IMG. j G j Place; S419, will add the specified motion image to the IMG j Add to the current intermediate image set; set j=j+1 and execute S414; the initial value of the current intermediate image set is empty; S420, the current intermediate image set is used as the target image set corresponding to the target video stream.

4. The method according to claim 2 or 3, characterized in that, Specify the center of the bounding box of the moving object as the corresponding addition point.

5. The method according to claim 2, characterized in that, Following S460 are the following steps: S462, for IMG j The target object in the image is identified to obtain the corresponding recognition box; the target object includes the specified moving object; S464, if DIMG j The bounding box and IMG corresponding to the specified moving object segmented from the image. j If there is an overlapping region between any two bounding boxes r in d, j <d r Occlude the region belonging to the recognition box r in the overlapping area. If d j >d r , will include DIMG in the overlapping region j Occlude the region of the bounding box corresponding to the specified moving object segmented in the middle, d r The distance between the frame r and the corresponding shooting device is given by j, which ranges from 1 to f(i)-1, where f(i) is the distance between the frame r and the shooting device. j The number of bounding boxes in the code; execute S470.

6. The method according to claim 3, characterized in that, Following S418 are the following steps: S4181, for IMG j The target object in the image is identified to obtain the corresponding recognition box; the target object includes the specified moving object; S4182, if DIMG j The bounding box and IMG corresponding to the specified moving object segmented from the image. j If there is an overlapping region between any two bounding boxes r in d, j <d r Occlude the region belonging to the recognition box r in the overlapping area. If d j >d r , will include DIMG in the overlapping region j Occlude the region of the bounding box corresponding to the specified moving object segmented in the middle, d r The distance between the frame r and the corresponding shooting device is given by j, which ranges from 1 to f(i)-1, where f(i) is the distance between the frame r and the shooting device. j The number of bounding boxes in the code; execute S419.

7. The method according to claim 1, characterized in that, The specified moving object is a pedestrian or a vehicle.

8. A device for expanding a visual model training dataset, characterized in that, The device includes: The first acquisition module is used to acquire the video stream to which the specified motion object needs to be added, and use it as the target video stream; The path generation module is used to generate a specified motion path in the scene corresponding to the target video stream, and set the attribute information corresponding to the specified motion path, the attribute information including motion time and motion form; The second acquisition module is used to acquire shooting information for each point on the specified motion path. The shooting information includes shooting angle and shooting distance. The shooting angle of a point is the angle between the line connecting the point to the corresponding shooting device and the shooting line of sight of the shooting device. The shooting distance of a point is the distance between the point and the corresponding shooting device. The module obtains a shooting information set C = {C1, C2, ..., C...} corresponding to the motion path. i , ..., C n }, the i-th captured information C in C i For the image capture information of the i-th point on the specified motion path, C i =(α i d i ), where i ranges from 1 to n, n is the number of points on the specified motion path, and α i Let d be the shooting angle corresponding to the i-th point. i The shooting distance corresponding to the i-th point; The addition module is used to obtain a two-dimensional image of the specified moving object at each point based on the image set of the target video stream to which a specified moving object needs to be added, based on C and the 3D model corresponding to the specified moving object, and to place the obtained two-dimensional image into the corresponding position of the corresponding image in the target video stream.

9. An electronic device, characterized in that, Including processor and memory; The processor executes the steps of the method as described in any one of claims 1 to 8 by invoking programs or instructions stored in the memory.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store a program or instructions that cause a computer to perform the steps of the method as described in any one of claims 1 to 8.

Images