Separated video generation method and handheld holder

Abstract

The invention relates to a split video generation method and a handheld holder, and the method comprises the steps: obtaining a plurality of continuous original video frame groups, including a key video frame group and a corresponding subsequent video frame group, calculating or updating splicing parameters through the key video frame group, and caching the splicing parameters, splicing a subsequent video frame group corresponding to the key video frame group by using the cached splicing parameters to obtain multiple split video frames; according to the method and the device for generating the duplicate video, the duplicate video can be quickly generated, and the calculation amount is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of video processing, and in particular to a method for generating a clone video and a handheld pan / tilt platform. Background Art

[0002] Avatar video is a creative special effects technology that uses technical means to allow the same person to appear in different positions, actions or identities in the same video screen at the same time, thus creating a refreshing visual effect. This special effects technology can be used in many scenarios, such as creative short videos: users can use "avatars" to simulate the interaction between different characters in the same scene, such as dialogue, cooperation, and even confrontation, to create a unique narrative effect. Teaching and demonstration: In educational videos or product demonstrations, users can use "avatars" to play multiple roles at the same time and intuitively demonstrate complex operations or processes. Entertainment and advertising: In entertainment videos or advertisements, the "avatar" special effect can attract the audience's attention, enhance the visual impact, and improve the dissemination and memory points of the work.

[0003] Green screen technology separates the foreground subject from the background by shooting a video with a monochrome background, and then superimposes and synthesizes multiple foregrounds. However, green screen technology requires green screen equipment and has complex layout.

[0004] The frame-by-frame stitching technology obtains multiple groups of original video frames by matching the video frames of multiple segments of original materials one by one, and then calculates the stitching parameters for each group of original video frames respectively, and finally uses the stitching parameters for stitching, which will result in a very large computational overhead. Summary of the invention

[0005] Aiming at the problem of how to quickly generate a clone video, the present invention proposes a clone video generation method and a handheld pan-tilt head.

[0006] In a first aspect, the present application provides a method for generating a clone video, the method comprising:

[0007] Acquire a plurality of continuous groups of original video frame groups, including a key video frame group and its corresponding subsequent video frame group, wherein the screen contents of the same group of original video frame groups include at least partially non-overlapping target objects;

[0008] Calculate or update splicing parameters using the key video frame group and cache them, wherein the splicing parameters are used to splice a group of original video frame groups into a clone video frame, wherein the clone video frame includes multiple target objects;

[0009] Using the cached splicing parameters to splice subsequent video frame groups corresponding to the key video frame groups to obtain multiple clone video frames; and

[0010] Generate a clone video based on multiple clone video frames.

[0011] In some embodiments, the clone video generation method is applied to a mobile terminal, which is mounted on a handheld gimbal. The multiple groups of original video frame groups include a continuous first to Nth group of original video frame groups. The method includes: using the first group of original video frame groups as a key video frame group; and using the second to Nth group of original video frame groups located after the key video frame group as subsequent video frame groups.

[0012] In some embodiments, the number of the key video frame groups is multiple groups, and the subsequent video frame groups corresponding to the key video frame groups are original video frame groups between the key video frame group and the next key video frame group.

[0013] In some embodiments, the using the cached splicing parameters to splice the subsequent video frame groups corresponding to the key video frame groups includes:

[0014] Before using the next key video frame group to update the splicing parameters, the corresponding subsequent video frame groups are spliced ​​using the splicing parameters calculated from the current key video frame group.

[0015] In some embodiments, the method further comprises: using the target parameter to identify the key video frame group and the subsequent video frame group.

[0016] In some embodiments, the target parameter includes a time interval; and the step of using the target parameter to confirm the key video frame group and the subsequent video frame group includes:

[0017] When the current original video frame group is separated from the previous key video frame group by a preset time interval, setting the current original video frame group as the key video frame group;

[0018] When the interval between the current original video frame group and the previous key video frame group is less than the preset time interval, the current original video frame group is set as the subsequent video frame group.

[0019] In some embodiments, the target parameter includes: a picture change feature value calculated based on the picture content of the previous group or multiple groups of original video groups; the step of using the target parameter to confirm the key video frame group and the subsequent video frame group includes:

[0020] When the picture change characteristic value is greater than a preset threshold, setting the current original video frame group as a key video frame group;

[0021] When the picture change characteristic value is less than or equal to a preset threshold, the current original video frame group is set as a subsequent video frame group.

[0022] In some embodiments, the using of the key video frame group to calculate or update the splicing parameters and cache them includes:

[0023] Based on the Akaze feature point detection algorithm, the brute force matching algorithm and the key video frame group, the splicing parameters between the original video frames of the key video frame group are calculated.

[0024] In some embodiments, the using the cached splicing parameters to splice the key video frame group and its corresponding subsequent video frame group respectively to obtain multiple clone video frames includes:

[0025] Using the homography matrix to perform image registration on the original video frames of the key video frame group, and confirming the initial pixel values ​​of the clone video frames;

[0026] Confirming the overlapping area between the original video frames of the key video frame group, and adjusting the initial pixel value of the overlapping area so that the brightness of the overlapping area transitions linearly;

[0027] Generate a clone video frame based on the initial pixel value corresponding to the cropped area.

[0028] In a second aspect, the present application provides a handheld gimbal, which is used to connect to a mobile terminal to control the camera direction of the mobile terminal; the mobile terminal is provided with a memory and a processor, the memory stores a clone video generation program, and the processor implements the above method when executing the clone video generation program.

[0029] The present application divides multiple continuous groups of original video frame groups into key video frame groups and subsequent video frame groups, and then only uses the key video frame groups to calculate splicing parameters, which are multiplexed as splicing parameters of the key video frame groups and subsequent video frame groups. With this arrangement, the number of calculations of the splicing parameters of the clone video generation method of the present application is far less than that of the traditional clone video generation method, thereby greatly reducing the amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a flowchart of an embodiment of a method for generating a cloned video according to the present invention;

[0031] Figure 2 It is a flowchart of another embodiment of the method for generating a cloned video of the present invention;

[0032] Figure 3 A schematic diagram of a group of original video frames in the method for generating a cloned video of the present invention;

[0033] Figure 4 A schematic diagram of a clone video frame obtained by splicing a group of original video frames in the clone video generation method of the present invention;

[0034] Figure 5 It is a schematic diagram of a clone video frame obtained by splicing a group of original video frames in the clone video generation method of the present invention after eliminating chromatic aberration. DETAILED DESCRIPTION

[0035] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0036] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0037] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0038] Example 1

[0039] It should be noted that the method of generating clone videos by calculating and splicing them frame by frame requires calculating a set of splicing parameters for each clone video frame. As the number of clone video frames increases, the amount of calculation increases linearly and the processing time is long. It is difficult to efficiently complete the splicing of dynamic video clones on ordinary mobile devices.

[0040] In response to the above problems, the present application provides a method for generating a clone video. The method realizes efficient splicing of dynamic video clones on ordinary mobile devices by reusing splicing parameters.

[0041] Reference Figure 1 In one embodiment, the method for generating a clone video includes:

[0042] S100, obtaining a plurality of continuous groups of original video frame groups, including a key video frame group and its corresponding subsequent video frame group, wherein the picture contents of the same group of original video frame groups include at least partially non-overlapping target objects. In addition, the picture contents of the same group of original video frame groups at least partially overlap to meet the splicing requirements.

[0043] In the embodiment of the present application, the execution subject can be a shooting device or a pan-tilt device. In addition, the clone video generation method of the present invention can be formed in application software and then installed in a mobile terminal.

[0044] The shooting device shoots video segments of the target object in different positions, different actions or playing different roles under a fixed background, ensuring that the scenes of multiple video segments are basically consistent, and the positions of the characters do not overlap or partially overlap, and then splicing the multiple video segments to obtain a clone video.

[0045] In this embodiment, in order to confirm the original video frame groups corresponding to multiple video segments, we can use parameters such as picture content, timestamp, frame number, etc. to match the corresponding video frames between multiple video segments to obtain multiple continuous groups of original video frame groups. Among them, the continuous multiple groups of original video frame groups refer to the playback order of the generated clone video frames. The specific video frame matching method is not limited here. For example, when shooting each video segment, a specific action / sound / marker can be used as a matching basis. After matching one or more groups of video frame groups, the timestamp can be used to match the video frames after this group or multiple groups of video frame groups to obtain multiple groups of original video frame groups.

[0046] In this embodiment, each group of original video frames is used to generate a frame of clone video frame. In order to achieve feature matching between original video frames, there must be overlapping areas between original video frames for feature matching. Feature matching of overlapping areas may include feature detection and feature point matching. In this embodiment, feature point detection is not limited. Exemplarily, feature point detection can use SIFT (Scale-Invariant Feature Transform): it can extract stable feature points under scale and rotation changes, which is suitable for matching needs of complex scenes. Although the accuracy is high, the calculation cost is relatively high and it is not suitable for real-time processing. ORB (Oriented FAST and Rotated BRIEF) can also be used: combining FAST feature point detection with BRIEF descriptor, the calculation speed is faster, while maintaining a certain matching accuracy. It is more suitable for application on mobile devices to meet the needs of resource-constrained scenarios. AKAZE (Accelerated-KAZE) can also be used: based on nonlinear scale space, it can extract high-quality feature points and has good calculation efficiency. It is suitable for feature point detection scenarios of dynamic videos.

[0047] The matching algorithm is not limited here. For example, the matching algorithm can use brute-force matching (Brute-Force Matcher), which compares the descriptors of the feature points one by one, calculates the similarity (such as Euclidean distance), and finds the best matching pair. It is simple and easy to implement, but the calculation speed is slow, and it is suitable for offline processing. You can also use FLANN (Fast Library for Approximate Nearest Neighbors): based on the approximate nearest neighbor search algorithm, it can quickly match a large number of feature points. It is more common in real-time applications, especially when there are many feature points, the efficiency is significantly improved. You can also use KNN (K-Nearest Neighbors): by selecting multiple candidate matching pairs and combining them with a ratio test (such as Lowe's ratio test) to screen out the best match. It can effectively reduce mismatches and improve the accuracy of splicing.

[0048] S200, using a key video frame group to calculate or update a splicing parameter and cache it, wherein the splicing parameter is used to splice a group of original video frames into a clone video frame, wherein the clone video frame includes a plurality of target objects;

[0049] It should be noted that the present application divides the original video frame group into a key video frame group and a subsequent video frame group, and only calculates the splicing parameters of the key video frame group, rather than using each group of original video frame groups to calculate the corresponding splicing parameters and then splicing them frame by frame. The splicing parameters obtained in the present application are not only used for the splicing of the key video frame group, but also reused for the splicing of the corresponding subsequent video frame groups. With such a setting, the number of calculations of the splicing parameters of the clone video generation method of the present application is far less than that of the traditional clone video generation method, which greatly reduces the amount of calculation.

[0050] S300, using the cached splicing parameters to splice subsequent video frame groups corresponding to the key video frame group to obtain multiple clone video frames; and

[0051] It should be noted that multiple groups of original video frames can be recorded as the first to Nth groups of original video frame groups. The key video frame group can be the first group of original video frame groups, or it can be the original video frame group corresponding to the moment when the splicing parameters need to be updated (for example, when the scene changes, or periodically updated).

[0052] S400: Generate a clone video according to multiple clone video frames, that is, encapsulate the obtained multiple clone video frames into a video file, such as an MP4 file.

[0053] The present application divides multiple continuous groups of original video frame groups into key video frame groups and subsequent video frame groups, and then only uses the key video frame groups to calculate splicing parameters, which are multiplexed as splicing parameters of the key video frame groups and subsequent video frame groups. With this arrangement, the number of calculations of the splicing parameters of the clone video generation method of the present application is far less than that of the traditional clone video generation method, thereby greatly reducing the amount of calculation.

[0054] In some embodiments, the plurality of groups of original video frame groups include a first group to an Nth group of continuous original video frame groups, and the method includes: using the first group of original video frame groups as a key video frame group; and using the second group to the Nth group of original video frame groups located after the key video frame group as subsequent video frame groups. In other words, the present application only needs to calculate the splicing parameters once to generate a cloned video.

[0055] It should be noted that the clone video generation method of the present application is applied to the combination of a handheld gimbal and a mobile terminal. The inventors have found through actual tests that this form of combined shooting will not move significantly, so the background will not change significantly. Only the first group of original video frames is used as the key video frame group, which can effectively reduce the amount of calculation while meeting the needs.

[0056] Of course, in other application scenarios, more key video frame groups need to be set. At this time, it is necessary to introduce target parameters as a reference to confirm the key video frame groups. For example, key video frame groups are set according to time or frame intervals. For example, the captured image is used to confirm whether the background has changed significantly, and whether to add key video frame groups.

[0057] In some embodiments, the number of the key video frame groups is multiple groups, and the subsequent video frame groups corresponding to the key video frame groups are original video frame groups between the key video frame group and the next key video frame group.

[0058] In other words, the key video frame group of this embodiment is not just one group, but multiple groups of key video frame groups, which are distributed in N groups of original video frame groups in a certain design method, so that the splicing effect of the splicing parameters can be ensured. The specific design method can be adjusted according to actual needs, and the key video frame group can be confirmed as needed to update the splicing parameters.

[0059] Specifically, the target parameters can be used to determine which of the N groups of original video frame groups are used as key video frame groups and which are used as subsequent video frame groups. Exemplarily, satisfying the target parameters can refer to reaching a preset interval time, reaching the number of original video frame groups at a preset interval, a significant change in the background / scene, a significant change in the captured image, etc. At this time, the original video frame groups that meet the target parameters are used as key video frame groups, and the original video frame groups that are located after the current key video frame group and before the next key video frame group are used as subsequent video frame groups of the current key video frame group.

[0060] Exemplarily, the target parameter may include a time interval; at this time, the target parameter is used to confirm the key video frame group and the subsequent video frame group, including: when the current original video frame group is separated from the previous key video frame group by a preset time interval, the current original video frame group is set as the key video frame group; when the interval between the current original video frame group and the previous key video frame group is less than the preset time interval, the current original video frame group is set as the subsequent video frame group.

[0061] In this embodiment, the processor processes in the order of N groups of original video frame groups, for example, the first group of original video frame groups is a key video frame group, the second to tenth groups of original video frame groups are subsequent video frame groups of the first group of original video frame groups, the eleventh group of original video frame groups is a key video frame group, and the twelfth to twentieth groups of original video frame groups are subsequent video frame groups of the eleventh group of original video frame groups.

[0062] In this embodiment, the first group of original video frame groups can be used to calculate splicing parameters and cache them, and the cached splicing parameters can be used to splice the first to tenth groups of original video frame groups respectively to obtain 10 clone video frames, and then the eleventh group of original video frame groups can be used to calculate splicing parameters and cache them (that is, the cached splicing parameters are updated), and then the updated splicing parameters can be used to splice the eleventh to twentieth groups of original video frame groups respectively to obtain 10 clone video frames, and so on, N groups of original video frame groups are spliced ​​into N frames of clone video frames, and then encapsulated into one clone video.

[0063] The above embodiment uses the number of interval original video frame groups as the target parameter to identify the key video frame groups. Similarly, the background / scene change amount and the shot picture change amount can also be used to identify the key video frame groups.

[0064] Exemplarily, the target parameters include: a picture change characteristic value calculated based on the picture content of a previous group or multiple groups of original video groups; the use of the target parameters to confirm the key video frame group and the subsequent video frame group includes: when the picture change characteristic value is greater than a preset threshold, setting the current original video frame group as the key video frame group; and when the picture change characteristic value is less than or equal to a preset threshold, setting the current original video frame group as the subsequent video frame group.

[0065] In this embodiment, the picture change characteristic value can be a characteristic value such as an optical flow, a pixel difference, or a mean or median of RGB pixels. For example, the scene change can be judged by the color change, and the average value of RGB (red, green, and blue) of the video frame is calculated. If it is greater than a threshold, it can be considered that a large change (scene change) has occurred in the picture. At this time, the current original video frame group is set as a key video frame group, and the subsequent original video frame groups are spliced. When it is detected that the average value of RGB (red, green, and blue) is greater than the threshold again, the current original video frame group is used as the key video frame group to update the splicing parameters for the subsequent splicing of the original video frame groups, and so on. Every time the scene changes, the splicing parameters are updated to better adapt to the situation where the scene changes frequently.

[0066] In some embodiments, the using cached stitching parameters to stitch subsequent video frame groups corresponding to the key video frame groups includes: before using the next key video frame group to update the stitching parameters, using the stitching parameters calculated for the current key video frame group to stitch the corresponding subsequent video frame groups.

[0067] In other words, the application generates a clone video in real time without too much waiting time. Specifically, the application can process frame by frame, use the cached splicing parameters to splice subsequent video frame groups, and when encountering a key video frame group, update the splicing parameters, and then continue to splice subsequent video frame groups, which can achieve streaming splicing and meet the dual needs of transmission and splicing.

[0068] In some embodiments, the use of key video frame groups to calculate or update stitching parameters and cache includes: based on an akaze feature point detection algorithm, a brute force matching algorithm and key video frame groups, calculating stitching parameters between original video frames of the key video frame groups.

[0069] Actual research has found that users use the scene more often indoors, where the light is usually darker and the background texture is simple. This embodiment uses the AKAZE feature point detection algorithm to generate high-quality feature points that are suitable for indoor environments. The Brute-Force Matcher is used for feature point matching, and the feature points with large deviations are filtered through the Hamming distance. The engineering implementation is simple, making the algorithm more robust.

[0070] Reference Figure 2 , a method for splicing an original video frame group is given. In this embodiment, the number of clones is three, that is, the clone video frame includes 3 target objects.

[0071] Reference Figure 3 , Figure 3The three original video frames of a group of original video frames are shown in an unjoined state. It can be seen that the three original video frames all have the target object (task), and the three original video frames all have overlapping areas. At this time, feature point matching can be performed through the overlapping area to stitch the three images into one image, as follows:

[0072] S301, using the homography matrix to perform image registration on the original video frames of the key video frame group, and confirming the initial pixel values ​​of the clone video frames;

[0073] Reference Figure 4 , we can use the middle picture as the reference, and align the left and right images relative to the middle, so that they are aligned to the same perspective. This step uses feature matching to calculate the relative homography matrix, such as the homography matrix of the left relative to the middle and the right relative to the middle. The homography matrix describes the relationship between the two images, including translation, rotation, scaling and perspective transformation. Figure 4 Stitched image shown.

[0074] S302, confirming the overlapping area between the original video frames of the key video frame group, and adjusting the initial pixel value of the overlapping area to make the brightness of the overlapping area transition linearly;

[0075] Reference Figure 4 , Figure 4 The stitched image shown is a picture of three people stitched together. However, the overlapping area is very abrupt. Therefore, we can calculate the overexposed overlapping area in the image and use linear transition to make the overlapping area look natural. This process can also be called the process of achromatic aberration. Specifically, the pixel brightness gain of the overlapping area can be calculated by the pixel value of the non-overlapping area. The brightness of the overlapping area can be adjusted using the brightness gain to eliminate the color difference (also called brightness difference). Figure 5 The specific achromatic aberration algorithm is not limited here.

[0076] S303: Generate a clone video frame based on the initial pixel value corresponding to the cropped area.

[0077] Reference Figure 5 , we can see that there are still a lot of black areas in the fused image. By calculating and cropping to get the useful area in the middle, the whole stitching process is completed.

[0078] It should be noted that in the above process, translation, rotation, scaling, perspective transformation, and cropping of the region can be achieved by rendering, that is, pixel movement. In other embodiments, it can also be achieved by splicing alignment and direct cropping.

[0079] Example 2

[0080] In a second aspect, the present application provides a handheld gimbal, which is used to connect to a mobile terminal to control the camera direction of the mobile terminal; the mobile terminal is provided with a memory and a processor, the memory stores a clone video generation program, and the processor implements the above method when executing the clone video generation program.

[0081] Example 3

[0082] In a third aspect, the present application provides a storage medium storing the above-mentioned clone video generation program. The clone video generation method of this embodiment can be formed into software, stored in a server, and downloaded and used by users. When a user downloads and runs the application, if the corresponding clone video generation function is selected, the above-mentioned clone video generation method can be implemented.

[0083] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0084] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0085] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0086] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0087] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0088] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0089] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A method for generating a clone video, characterized in that: include: Acquire a plurality of continuous groups of original video frame groups, including a key video frame group and its corresponding subsequent video frame group, wherein the screen contents of the same group of original video frame groups include at least partially non-overlapping target objects; Calculate or update splicing parameters using the key video frame group and cache them, wherein the splicing parameters are used to splice a group of original video frame groups into a clone video frame, wherein the clone video frame includes multiple target objects; Using the cached splicing parameters to splice subsequent video frame groups corresponding to the key video frame groups to obtain multiple clone video frames; and Generate a clone video based on multiple clone video frames.

2. The method for generating a clone video according to claim 1, wherein: The method for generating a cloned video is applied to a mobile terminal, the mobile terminal is mounted on a handheld gimbal, and the multiple groups of original video frame groups include a first group to an Nth group of continuous original video frame groups. The method includes: Using the first group of original video frames as a key video frame group; and The second to Nth groups of original video frame groups located after the key video frame group are taken as subsequent video frame groups.

3. The method for generating a cloned video according to claim 1, wherein: The number of the key video frame groups is multiple, and the subsequent video frame groups corresponding to the key video frame groups are the original video frame groups between the key video frame group and the next key video frame group.

4. The method for generating a clone video according to claim 3, wherein: The step of using the cached splicing parameters to splice subsequent video frame groups corresponding to the key video frame groups includes: Before using the next key video frame group to update the splicing parameters, the corresponding subsequent video frame groups are spliced ​​using the splicing parameters calculated from the current key video frame group.

5. The method for generating a cloned video according to claim 3 or 4, wherein: The method further includes: using the target parameter to identify a key video frame group and a subsequent video frame group.

6. The method for generating a clone video according to claim 5, wherein: The target parameter includes a time interval; the step of using the target parameter to confirm a key video frame group and a subsequent video frame group includes: When the current original video frame group is separated from the previous key video frame group by a preset time interval, setting the current original video frame group as the key video frame group; When the interval between the current original video frame group and the previous key video frame group is less than the preset time interval, the current original video frame group is set as the subsequent video frame group.

7. The method for generating a clone video according to claim 5, wherein: The target parameter includes: a picture change feature value calculated based on the picture content of the previous group or multiple groups of original video groups; the use of the target parameter to confirm the key video frame group and the subsequent video frame group includes: When the picture change characteristic value is greater than a preset threshold, setting the current original video frame group as a key video frame group; When the picture change characteristic value is less than or equal to a preset threshold, the current original video frame group is set as a subsequent video frame group.

8. The method for generating a cloned video according to claim 1, wherein: The method of using the key video frame group to calculate or update the splicing parameters and cache them includes: Based on the Akaze feature point detection algorithm, the brute force matching algorithm and the key video frame group, the splicing parameters between the original video frames of the key video frame group are calculated.

9. The method for generating a clone video according to claim 8, wherein: The step of using the cached splicing parameters to splice the key video frame group and its corresponding subsequent video frame group to obtain multiple clone video frames includes: Using the homography matrix to perform image registration on the original video frames of the key video frame group, and confirming the initial pixel values ​​of the clone video frames; Confirming the overlapping area between the original video frames of the key video frame group, and adjusting the initial pixel value of the overlapping area so that the brightness of the overlapping area transitions linearly; Generate a clone video frame based on the initial pixel value corresponding to the cropped area.

10. A handheld gimbal, characterized in that: The handheld gimbal is used to connect to a mobile terminal to control the camera direction of the mobile terminal; The mobile terminal is provided with a memory and a processor, the memory stores a clone video generation program, and when the processor executes the clone video generation program, the method according to any one of claims 1-9 is implemented.

Images