Intelligent video automatic generation method and system based on chapter random combination

Abstract

The invention discloses an intelligent video automatic generation method and system based on chapter random combination, and belongs to the technical field of video content automatic generation. The method comprises the steps that chapter parameters configured by a user are received, and a corresponding combination number calculation algorithm is selected according to a chapter mode to obtain the number of videos capable of being generated; generating a timeline based on the parameters and the number, wherein the process comprises material selection, sound and picture alignment and special effect parameter generation; the standardized timeline file is exported, an automatic script is generated, and batch rendering is executed and completed in professional software; the core lies in supporting five chapter modes, double-strategy audio and video alignment and a multi-round material extraction mechanism, and the repetition rate can be accurately controlled. According to the method, video batch automatic production is realized, the efficiency and content uniqueness are improved, the audio and picture synchronization precision is guaranteed, multi-scene and cross-platform use is adapted, and professional creation requirements are met.

Description

Technical Field

[0001] This invention belongs to the field of automatic video content generation technology, specifically relating to an intelligent video automatic generation method and system based on random combination of chapters. Background Technology

[0002] In the current field of mass video creation, traditional methods still rely on manual segment-by-segment editing: creators need to manually select materials, adjust their order, and match audio and video. Not only is the production of a single video time-consuming, but mass production is also prone to quality fluctuations due to fatigue or operational differences. While existing tools can achieve simple material splicing, they are mostly limited to applying fixed templates and lack intelligent combination logic at the chapter level. For example, they cannot flexibly combine materials according to actual needs such as "one voiceover paired with multiple videos" or "video looping to fill the voiceover duration," nor can they accurately control the repetition rate of materials. This results in highly similar video content generated from the same materials, which is easily identified as duplicate content by the platform and thus subject to traffic restrictions. In addition, the matching method between voiceover and video duration is also very simple, mostly relying on simple cropping or stretching of videos to adjust the duration, which makes it difficult to achieve high-precision audio-visual synchronization and cannot meet the needs of professional creation. Specifically, the following technical problems are raised: Mass production is inefficient and of inconsistent quality: Traditional manual editing requires processing the footage segment by segment. Mass production is time-consuming and labor-intensive, and the quality of videos in the same batch is easily inconsistent due to differences in human operation. Difficulty in controlling material duplication rate: Existing tools lack a chapter-level random combination mechanism, resulting in high similarity of video content generated based on the same materials, which can easily trigger platform traffic restrictions and affect the content dissemination effect; Insufficient audio-visual synchronization accuracy: The matching method between dubbing and video duration is simplistic, relying solely on cropping or stretching to adjust the duration. This approach cannot flexibly adapt to the characteristics of the material, resulting in large synchronization errors and failing to meet the stringent duration matching requirements of professional creation. To address this, we propose an intelligent automatic video generation method and system based on random chapter combinations. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent video automatic generation method and system based on random combination of chapters, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an intelligent video automatic generation method based on random chapter combination, comprising: Step 1: Receive the chapter parameters configured by the user. The chapter parameters include: the number of video materials, the number of audio materials, the number of randomly selected videos for each voiceover, the repetition rate control level, the voiceover reuse mode, and the chapter mode. Step 2: Select the corresponding combination number calculation algorithm according to the chapter mode, and obtain the number of video combinations that can be generated through the combination number calculation algorithm; Step 3: Based on the chapter parameters and the number of video combinations, generate a specified number of timelines. The process of generating timelines includes: traversing all chapter configurations, selecting audio materials according to the dubbing reuse mode, selecting video materials through a multi-round material extraction mechanism, calculating audio-visual duration alignment parameters, and generating video effects parameters. Step 4: Export the generated timeline as an Open Timeline Exchange Format (OTF) standard file; Step 5: Based on the Open Timeline Exchange Format standard file and material path information, generate a script for automated execution by professional video editing software; Step Six: Execute the script in professional video editing software to complete the batch rendering and export of videos.

[0005] Preferably, it is characterized by: The chapter mode includes five distinct types: one voice-over with multiple videos, fill-in-the-screen mode, loop plot mode, no voice-over with fixed number of materials mode, and no voice-over with specified duration mode. The "one voice-over, multiple video" mode refers to a combination mode in which a voice-over is randomly matched with multiple video clips. The screen-filling mode refers to a combination mode that uses video clips to fill the screen based on the dubbing duration. The looping plot mode refers to a combination mode that uses sub-chapter video clips to play in a loop to match the dubbing duration; The "no-dubbing fixed-materials-number-number-mode" refers to a combination mode that requires no dubbing and selects a fixed number of video clips. The "no voiceover specified duration mode" refers to a combination mode in which no voiceover is required and the total video duration must be filled to the preset target duration.

[0006] Preferably, it is characterized by: The dubbing reuse mode includes two explicitly selectable types: single-use mode and repeatable mode. The single-use mode means that the same voiceover can only be used to generate one video timeline, while the repeatable mode means that the same voiceover can be used to generate multiple video timelines. The process of selecting audio materials according to the dubbing reuse mode is as follows: if it is a single-use mode, then unused dubbings are selected from the audio material library; if it is a repetition allowed mode, then any dubbing is randomly selected from the audio material library, without restricting whether the dubbing has been used.

[0007] Preferably, it is characterized by: The specific logic of the combination number calculation algorithm selected according to the chapter mode is as follows: if the chapter mode is a one-dubbing-multiple-video mode, then the permutation algorithm is used to calculate the combination number; If the chapter mode is a fill screen mode, then the minimum or maximum value algorithm is selected to calculate the number of combinations based on the dubbing reuse mode; If the chapter mode is a loop story mode, then a calculation algorithm based on the number of sub-chapter videos and the loop count will be used; If the chapter mode is a fixed number of materials without voice-over, then a combination algorithm based on the number of video materials and the fixed number of selected materials will be used; If the chapter mode is a no-dubbing, specified duration mode, then a calculation algorithm based on the average video duration and the target duration will be used.

[0008] Preferably, it is characterized by: The specific execution logic of the multi-round material extraction mechanism is as follows: using the user-configured repetition rate control level as the initial extraction condition, video segments are extracted from the video material library; After extraction is complete, calculate the total duration of all extracted segments. If the total duration reaches the preset target duration, the extraction process ends. If the total duration does not reach the preset target duration and the current repetition rate control level does not reach the highest level, the repetition rate control level will be increased by one level to relax the restrictions on material reuse, and the extraction will be performed again. Repeat the above process of extraction, duration check, and gear adjustment until the total duration of the extracted segments reaches the target or the preset maximum number of extraction rounds is reached, in order to avoid infinite loops.

[0009] Preferably, it is characterized by: The calculation logic for the audio-visual duration alignment parameter includes two selectable strategies: a cutting strategy and a speed-changing strategy. The cutting strategy refers to cutting out a portion of a video clip to match the target duration. The calculation parameters include the cutting start offset and the cutting duration. The speed-adjusting strategy refers to adjusting the playback speed of video segments to match the target duration. The calculated parameter is a speed coefficient, which is equal to the ratio of the total duration of the extracted video segments to the target duration. The choice between the two strategies is either preset by the user based on the characteristics of the material, or automatically matched by the system based on the requirements for the integrity of the video content.

[0010] Preferably, it is characterized by: The video effects parameters include three categories of explicitly configurable parameters: mirror parameters, cropping parameters, and scaling parameters. The mirroring parameter is used to control whether horizontal or vertical mirroring effects are added to the video, and the configuration option is to enable or disable it. The cropping parameters are used to control the cropping range of the video frame, and the configuration content is the ratio or pixel range of the cropping area; The scaling parameter is used to control the scaling ratio of the video image, and the configuration content is a scaling factor or a scaling ratio corresponding to the target resolution.

[0011] Preferably, it is characterized by: The specific data structure of the Open Timeline Exchange Format standard file includes video track data units and audio track data units; The video track data unit contains the media reference path, time range information, and video effects parameters for each video segment; The media reference path points to the material storage location, the time range information includes the start time and duration of the segment in the timeline, and the video effect parameters are associated with generated mirror parameters, cropping parameters, and scaling parameters. The audio track data unit contains the media reference path, time range information, and volume control parameters for each audio segment. The time range information is aligned with the time range of the corresponding video segment, and the volume control parameters are used to adjust the volume of the dubbing or background music.

[0012] Preferably, it is characterized by: The core components of the script used for automated execution by professional video editing software include a material import command unit, a timeline import command unit, a rendering export command unit, and a path processing unit. The material import command unit is used to instruct professional video editing software to import audio and video materials into the media pool in batches; The timeline import command unit is used to instruct professional video editing software to read the Open Timeline Exchange Format standard file and automatically construct the corresponding video timeline; The rendering export command unit is used to instruct professional video editing software to configure rendering parameters and add rendering tasks to the queue. The rendering parameters include resolution and encoding format. The path processing unit is used to adapt to file path formats of different operating systems, including compatibility processing for long paths and automatic escaping of path separators, to ensure that the script can be executed normally in different system environments.

[0013] Compared with the prior art, the beneficial effects of the present invention are: (1) This intelligent video automatic generation method and system based on random combination of chapters, by relying on the full-process automated design of "parameter configuration - combination calculation - timeline generation - script rendering", combined with cross-platform adaptation and standardized format, realizes batch production from raw materials to finished video without human intervention; no need for manual editing, one-by-one alignment of audio and video or configuration of rendering parameters, completely solving the pain points of traditional manual editing that are "time-consuming, cumbersome and quality fluctuating", and at the same time, through multi-threaded concurrent calculation and batch rendering mechanism, the large-scale video production cycle is greatly shortened, which is suitable for the "high efficiency and large batch" creation needs of MCN agencies, educational institutions and other organizations.

[0014] (2) This intelligent video automatic generation method and system based on chapter random combination generates a large amount of differentiated content from limited materials through chapter-level random combination algorithm, multi-round material extraction mechanism and 5-level repetition rate control, effectively avoiding platform traffic limitation caused by material repetition; at the same time, it innovatively adopts a dual strategy of cutting and speed adjustment to achieve audio-visual duration alignment, and with quantitative calculation of speed coefficient and cutting parameters, ensures that the audio-visual synchronization error is controlled at the professional level, which not only ensures the integrity of the content (such as the absence of problem-solving steps in teaching videos), but also meets the requirements of professional creation for audio-visual accuracy.

[0015] (3) The intelligent video automatic generation method and system based on random combination of chapters supports five chapter modes and two dubbing reuse modes, which can flexibly adapt to different creation scenarios such as "one dubbing for multiple videos" and "fixed duration without dubbing". The system is compatible with mainstream operating systems and professional video editing software. Through the visual parameter configuration interface and automated scripts, the operation threshold for non-technical personnel is reduced, and batch creation can be completed without complicated learning. At the same time, the standardized file format and path adaptation logic ensure the stability of cross-tool and cross-platform use and broaden the scope of technology application. Attached Figure Description

[0016] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1; Please see Figure 1 Figure 1 illustrates an intelligent video automatic generation method based on random combination of chapters, comprising: Step 1: Operators configure chapter parameters Educational institution operators configure chapter parameters through a front-end interactive interface. The configuration process revolves around the structuring requirements of knowledge point videos, and all core parameters are defined with symbols to support subsequent algorithm calculations, as detailed below: Material quantity configuration: Upload pre-recorded example demonstration videos and imported knowledge point voiceovers from your institution. The system will automatically identify and count the total quantity of both types of materials, defined as follows: V: Total number of video materials (i.e., the total number of example demonstration videos); A: Total number of audio materials (i.e., the total number of knowledge points imported into the dubbing); Video structure configuration: Select the "1 voice-over + several videos" mode from the video structure options in the interface, which is called the chapter mode M=1, and set the "random number of videos to be selected for each voice-over", which is defined as n (in line with the teaching logic of "import + example", such as n corresponding to "1 voice-over paired with n example videos"). Repeatability control configuration: Adjust the repeatability control level using the slider, and define the repeatability control level as R; where the repeatability control level R ranges from 0 to 4, where 0 represents the strictest repeatability control and 4 represents the most lenient repeatability control. In this scenario, the middle level is selected to balance content quality and generation quantity. Voiceover reuse configuration: Select the voiceover reuse mode, denoted as S, where S=0 indicates single use and S=1 indicates that repetition is allowed. In this scenario, S=1 is set to maximize the use of high-quality voiceovers. After the parameters are configured, the operator clicks the "Submit" button. The system then transmits the total number of video materials (V), the total number of audio materials (A), the number of randomly selected videos for each dubbing (n), the repetition rate control level (R), the dubbing reuse mode (S), and the chapter mode (M) to the core algorithm module via encrypted communication. It should be noted that this step achieves quantitative configuration through "binding Chinese names and symbols of parameters," completely solving the pain point of "parameter ambiguity" in traditional tools. In the past, operators could only subjectively choose "more repetition" or "less repetition," and could not accurately control the content structure. However, this solution uses quantitative parameters such as the repetition rate control level R and the number of videos randomly selected for each dubbing n to make the repetition degree and video structure accurately defined, avoiding the quality fluctuation of the same batch of videos caused by parameter ambiguity.

[0019] Step 2: The core algorithm calculates the number of video combinations that can be generated; The core algorithm module is based on the parameters defined in step 1: chapter mode M, total number of video materials V, total number of audio materials A, number of randomly selected videos for each dubbing n, repetition rate control level R, and dubbing reuse mode S; the logic for calculating the number of combinations according to the chapter mode is as follows: Algorithm selection logic: Since the chapter mode M=1 (1 voice-over + several videos), the voice-over reuse mode S=1 (repetition is allowed), and the repetition rate level (R<4) (not reaching the highest level), the permutation algorithm is used to calculate the number of unique video combinations that can be generated, denoted as P(V,n). Permutation Algorithm Formula and Character Definition: The permutation algorithm is used to calculate the total number of combinations of "randomly selecting n videos from V video clips, without considering the selection order but allowing duplicate clips (since S=1)". The formula is: ; Where: P(V,n): the number of unique video combinations that can be generated; V: the total number of video materials (defined in step 1); n: the number of randomly selected videos corresponding to each dubbing (defined in step 1). Combination limit check: To avoid "combination explosion" (i.e., P(V,n) far exceeding actual production needs, resulting in waste of system resources) caused by an excessively large total amount of materials V, the algorithm presets a combination limit. ,like Then the number of combinations will be automatically adjusted to ; Output results: After the calculation is complete, the front-end interface displays the final number of videos that can be generated: Operator confirmation Once the video supply requirements of the online course platform are met, the timeline generation process is triggered.

[0020] It should be noted that this step addresses the pain point of "uncontrollable output" in traditional manual editing through a "quantitative algorithm"—previously, operators could only estimate output based on experience, often resulting in problems such as "insufficient output requiring temporary reshoots" or "excessive output causing equipment lag"; while this solution accurately calculates P(V,n) through a permutation algorithm, combined with an upper limit. The protection system ensures that mass production is "controllable in scale" and "plannable in pace," for example, when V is large. This can prevent the algorithm from getting bogged down in meaningless calculations of an extremely large number of data points, ensuring the continuity of the process.

[0021] Step 3: The core algorithm generates the video timeline; The core algorithm module is based on (Number of videos that can be generated) and the parameters from step 1 are used to generate independent video timelines one by one. During the generation process, the audio-visual alignment and material extraction logic are quantified using formulas, as detailed below: Timeline structure initialization: In M=1 mode, allocate one audio track (for storing dubbing) and one video track (for storing n topic videos) to each timeline, and define the target duration for each timeline. (That is, the selected dubbing duration is obtained automatically by the system from the dubbing file attributes). Audio material selection: Since S = 1 (allowing dubbing repetition), randomly select 1 piece from A dubbings and record its duration as ; Multi-round video material extraction: Initial extraction: Extract n pieces from V videos according to the repetition rate range R, and calculate the total duration of the extracted videos (that is, the sum of the durations of n videos, , where ti is the duration of the i-th video); Duration check and repetition rate adjustment: If , the extraction ends; if and (R < 4), then increase R by 1 level (relax the repetition control) and extract again until or the maximum extraction round is reached (to avoid infinite loop); Audio-visual duration alignment calculation: Since the physics example video needs to fully display the problem-solving steps, select a variable speed strategy and define the speed coefficient SpeedFactor. The formula is: ; Among them: SpeedFactor: The video playback speed coefficient (unitless). If (SpeedFactor > 1), it means the video needs to be accelerated (the total duration is shortened to ); if (0 < SpeedFactor < 1), it means the video needs to be decelerated (the total duration is extended to ); Ttotal: The total duration of the extracted videos (defined in step 3.3); : The timeline target duration (defined in step 3.1); For example: If is slightly greater than , then SpeedFactor is slightly greater than 1, and the video is slightly accelerated so that the duration is exactly aligned with the dubbing without affecting the integrity of the problem-solving steps; Video special effect parameter generation: According to the requirements of the online course platform, generate three types of quantitative special effect parameters: Mirror parameter: Set to "disabled" (to avoid confusion in the direction of physical experiment operations); Cropping ratio: Set to (16:9) (to adapt to multiple terminals); Scaling coefficient: Defined as ScaleFactor, and the formula is: ; Among them, Resoutput is the output resolution and Ressource is the original video resolution, ensuring that the screen is full-screen without black edges.

[0022] It should be noted that in this step, through the total duration of the extracted videos The SpeedFactor and ScaleFactor parameters are quantified to address the pain points of traditional manual editing, such as "large audio-visual alignment errors" and "subjective special effects parameters." When adjusting manually, the audio-visual synchronization error often exceeds 0.5 seconds, and special effects (such as scaling and cropping) rely on personal experience. In contrast, this solution accurately calculates the speed factor and scaling ratio using formulas, and the audio-visual synchronization error can be controlled within 0.01 seconds. The special effects parameters are uniformly adapted to platform requirements, eliminating the need for secondary adjustments.

[0023] Step 4: Export standardized timeline files using the core algorithm; The core algorithm module exports each timeline as a standardized OpenTimelineIO (OTIO) file. The file data structure is defined by quantization parameters to ensure compatibility with professional editing software, as detailed below: Video Track Data Unit: Media Reference Path: Records the storage path of the n title videos; Time Range Parameter: Defines the start time of each video segment. i and duration ,i; in: (The start time of the i-th video is equal to the sum of the durations of the previous (i-1) videos). (The duration of the i-th video = the original duration ti × the speed factor SpeedFactor); Effect parameters: Associate the ScaleFactor generated in step 3.5 with the cropping ratio (16:9). Audio track data unit: Media reference path: Records the storage path of the selected dubbing; Time range parameter: Start time =0 (starts synchronized with the video track), duration = (Same as the target duration); Volume parameter: Define the volume factor (VolFactor=0.9) (no unit), that is, the dubbing volume = original volume × VolFactor, to avoid drowning out the example explanation sound.

[0024] All OTIO files are named according to "generation date + timeline number" (e.g., "202412_TL_001.otio") and stored in designated folders for easy access by subsequent rendering scripts.

[0025] It should be noted that this step is based on the start time. i. Duration i. The VolFactor quantization parameter ensures the "standardization" and "accuracy" of the OTIO file. Traditional non-standard timeline files often have problems such as "segment misalignment" and "parameter loss". However, this solution defines the time range and attributes of each segment through formulas. After importing into different editing software (such as DaVinciResolve and Premiere), the timeline structure can be completely restored without manual correction.

[0026] Step 5: The core algorithm generates an automated rendering script; The core algorithm module generates an automated rendering script for professional editing software based on the OTIO file path and the material path. Key rendering parameters in the script are quantified and calculated using formulas, as follows: Material Import Command Unit: Instructs the software to batch import V video clips and A audio clips, automatically checks the validity of the material paths, and generates a log message if the path is invalid; Timeline Import Command Unit: Instructs the software to read all OTIO files and automatically build... Multiple timelines, synchronously applying parameters such as (SpeedFactor, ScaleFactor); Rendering parameter configuration command unit: Resolution: Set to =W×H (W is width, H is height, adapting to the general requirements of online learning platforms); Encoding format: Select H.264 (compatible with multiple terminals); Bitrate calculation: Define the target bitrate, the formula is: ; Where: Bitrate: target bitrate of video (unit: bps); k: bitrate coefficient (empirical value, no unit, value is taken according to the complexity of video, such as a larger k for dynamic scenes and a smaller k for static scenes). Output resolution (unit: pixels) =W×H); Fps: Output frame rate (unit: fps, such as 25fps, 30fps, depending on platform requirements); Output path and naming: Specify the output folder, and name the video according to "knowledge point type-serial number" (such as "Newton's First Law_001.mp4"). Path processing unit: Automatically adapts to the operating system's path format (such as escape path separators in Windows system) to ensure that the script is executable across platforms.

[0027] After the script is generated, it is stored in Lua format (e.g., “Render_Script.lua”). The front-end interface displays the script path and execution instructions (e.g., “Execute dofile('script path') in the editing software console”).

[0028] It should be noted that this step quantifies the bitrate using the target bitrate formula, solving the pain point of "blind bitrate setting" in traditional manual rendering. When manually setting the bitrate, problems often arise such as "too high a bitrate leading to an excessively large file size" or "too low a bitrate leading to blurry image quality." This solution, however, calculates the bitrate by combining video resolution, frame rate, and complexity using a formula, ensuring that the image quality meets the standards and the file size is adapted to the loading speed requirements of online learning platforms.

[0029] Step 6: The operator executes the script to complete batch rendering; The operator executes the rendering script in professional video editing software. The entire process is automated based on the quantization parameters in step 5, as detailed below: Script execution: Open the editing software, create a new project, enter the "Console", paste the script execution command and run it. The software will automatically trigger the following process: Material import: Import V videos and A audios in batches according to the script instructions. After the import is complete, the message "Materials ready (total (V+A) items)" will be displayed. Timeline build: Reads all OTIO files and builds automatically. There are multiple timelines, and operators can randomly check them—for example, to see if the SpeedFactor is working on a particular timeline. Is i accurate (the segment is not misaligned)? Adding to the rendering queue: Configure the rendering task by parameters such as (Bitrate, Res_{output}), and then... Add multiple timelines to the rendering queue in batches; Batch rendering: When the operator clicks "Start Rendering", the software automatically renders according to the queue order, and the rendering progress is displayed in real time. During rendering, if a video is interrupted due to material issues, the script will skip that video and log the process to prevent the overall workflow from terminating. Result verification: After rendering, the video is stored according to a preset path and name, and the system automatically verifies key parameters (such as whether the duration is correct). Is the resolution [missing information]? Once verified, it can be uploaded to the online course platform.

[0030] It should be noted that this step achieves "high efficiency" and "high quality" in batch rendering through "full-process quantitative parameter driving" - traditional manual rendering of a single video requires manually configuring more than 10 parameters, and Pfinal requires several hours of operation for a single video; while this solution is executed automatically by script, and batch rendering can be started in just a few minutes of configuration. Moreover, because the parameters are all quantified and calculated by formula, the consistency of the rendering results (such as image quality and duration) is far higher than that of manual operation, which greatly reduces the cost of post-processing correction.

[0031] An intelligent video automatic generation system based on random chapter combination includes: a user interface module, a communication middleware layer, a core algorithm module, and a professional rendering module; The user interface module is used to provide a visual operation interface, receive chapter parameter configuration input by the user, which corresponds to the chapter parameters in step 1 of the method, and simultaneously display the number of video combinations calculated by the core algorithm module in real time, which corresponds to the number of video combinations in step 2 of the method. The communication middleware layer is used to establish a communication link between the user interface module and the core algorithm module, realize the bidirectional transmission of chapter parameters, calculation instructions and result data, and manage file system access permissions to ensure the read and write security of materials and generated files. The core algorithm module is used to execute the core logic of steps two to five of the method. Specifically, it includes calling the corresponding combination number calculation algorithm according to the chapter mode, generating a timeline containing audio-visual alignment parameters and video effect parameters, exporting an open timeline exchange format standard file, and generating an automated script for professional video editing software. The professional rendering module is used to interface with mainstream professional video editing software, receive automated scripts generated by the core algorithm module, and execute the material import, timeline construction, and batch rendering commands in the scripts. This process corresponds to step six of the method, and finally completes the automated export of the video.

[0032] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for automatically generating intelligent videos based on random combination of chapters, characterized in that, This includes the following processes: Step 1: Receive the chapter parameters configured by the user. The chapter parameters include: the number of video materials, the number of audio materials, the number of randomly selected videos for each voiceover, the repetition rate control level, the voiceover reuse mode, and the chapter mode. Step 2: Select the corresponding combination number calculation algorithm according to the chapter mode, and obtain the number of video combinations that can be generated through the combination number calculation algorithm; Step 3: Based on the chapter parameters and the number of video combinations, generate a specified number of timelines. The process of generating timelines includes: traversing all chapter configurations, selecting audio materials according to the dubbing reuse mode, selecting video materials through a multi-round material extraction mechanism, calculating audio-visual duration alignment parameters, and generating video effects parameters. Step 4: Export the generated timeline as an Open Timeline Exchange Format (OTF) standard file; Step 5: Based on the Open Timeline Exchange Format standard file and material path information, generate a script for automated execution by professional video editing software; Step Six: Execute the script in professional video editing software to complete the batch rendering and export of videos.

2. The intelligent video automatic generation method based on random chapter combination according to claim 1, characterized in that: The chapter mode includes five distinct types: one voice-over and multiple video mode, fill-in-the-screen mode, loop plot mode, no voice-over and fixed number of materials mode, and no voice-over and specified duration mode. The "one voice-over, multiple video" mode refers to a combination mode in which a voice-over is randomly matched with multiple video clips. The screen-filling mode refers to a combination mode that uses video clips to fill the screen based on the dubbing duration. The looping plot mode refers to a combination mode that uses sub-chapter video clips to play in a loop to match the dubbing duration; The "no-dubbing fixed-materials-number-number-mode" refers to a combination mode that does not require dubbing and selects a fixed number of video clips. The "no voiceover specified duration mode" refers to a combination mode in which no voiceover is required and the total video duration must be filled to the preset target duration.

3. The intelligent video automatic generation method based on random chapter combination according to claim 2, characterized in that: The dubbing reuse mode includes two explicitly selectable types: single-use mode and repeatable mode. The single-use mode means that the same voiceover can only be used to generate one video timeline, while the repeatable mode means that the same voiceover can be used to generate multiple video timelines. The process of selecting audio materials according to the dubbing reuse mode is as follows: if it is a single-use mode, then unused dubbings are selected from the audio material library; if it is a repetition allowed mode, then any dubbing is randomly selected from the audio material library, without restricting whether the dubbing has been used.

4. The intelligent video automatic generation method based on random chapter combination according to claim 3, characterized in that: The specific logic of the combination number calculation algorithm selected according to the chapter mode is as follows: if the chapter mode is a one-dubbing-multiple-video mode, then the permutation algorithm is used to calculate the combination number; If the chapter mode is a fill screen mode, then the minimum or maximum value algorithm is selected to calculate the number of combinations based on the dubbing reuse mode; If the chapter mode is a loop story mode, then a calculation algorithm based on the number of sub-chapter videos and the loop count will be used; If the chapter mode is a fixed number of materials without voice-over, then a combination algorithm based on the number of video materials and the fixed number of selected materials will be used; If the chapter mode is a no-dubbing, specified duration mode, then a calculation algorithm based on the average video duration and the target duration will be used.

5. The intelligent video automatic generation method based on random chapter combination according to claim 4, characterized in that: The specific execution logic of the multi-round material extraction mechanism is as follows: using the user-configured repetition rate control level as the initial extraction condition, video segments are extracted from the video material library; After extraction is complete, calculate the total duration of all extracted segments. If the total duration reaches the preset target duration, the extraction process ends. If the total duration does not reach the preset target duration and the current repetition rate control level does not reach the highest level, the repetition rate control level will be increased by one level to relax the restrictions on material reuse, and the extraction will be performed again. Repeat the above process of extraction, duration check, and gear adjustment until the total duration of the extracted segments reaches the target or the preset maximum number of extraction rounds is reached, in order to avoid infinite loops.

6. The intelligent video automatic generation method based on random chapter combination according to claim 5, characterized in that: The calculation logic for the audio-visual duration alignment parameter includes two selectable strategies: a cutting strategy and a speed-changing strategy. The cutting strategy refers to cutting out a portion of a video clip to match the target duration. The calculation parameters include the cutting start offset and the cutting duration. The speed-adjusting strategy refers to adjusting the playback speed of video segments to match the target duration. The calculated parameter is a speed coefficient, which is equal to the ratio of the total duration of the extracted video segments to the target duration. The choice between the two strategies is either preset by the user based on the characteristics of the material, or automatically matched by the system based on the requirements for the integrity of the video content.

7. The intelligent video automatic generation method based on random chapter combination according to claim 6, characterized in that: The video effects parameters include three categories of explicitly configurable parameters: mirror parameters, cropping parameters, and scaling parameters. The mirroring parameter is used to control whether horizontal or vertical mirroring effects are added to the video, and the configuration option is to enable or disable it. The cropping parameters are used to control the cropping range of the video frame, and the configuration content is the ratio or pixel range of the cropping area; The scaling parameter is used to control the scaling ratio of the video image, and the configuration content is a scaling factor or a scaling ratio corresponding to the target resolution.

8. The intelligent video automatic generation method based on random chapter combination according to claim 7, characterized in that: The specific data structure of the Open Timeline Exchange Format standard file includes video track data units and audio track data units; The video track data unit contains the media reference path, time range information, and video effects parameters for each video segment; The media reference path points to the material storage location, the time range information includes the start time and duration of the segment in the timeline, and the video effect parameters are associated with generated mirror parameters, cropping parameters, and scaling parameters. The audio track data unit contains the media reference path, time range information, and volume control parameters for each audio segment. The time range information is aligned with the time range of the corresponding video segment, and the volume control parameters are used to adjust the volume of the dubbing or background music.

9. The intelligent video automatic generation method and system based on random chapter combination according to claim 8, characterized in that: The core components of the script used for automated execution by professional video editing software include a material import command unit, a timeline import command unit, a rendering export command unit, and a path processing unit. The material import command unit is used to instruct professional video editing software to import audio and video materials into the media pool in batches; The timeline import command unit is used to instruct professional video editing software to read the Open Timeline Exchange Format standard file and automatically construct the corresponding video timeline; The rendering export command unit is used to instruct professional video editing software to configure rendering parameters and add rendering tasks to the queue. The rendering parameters include resolution and encoding format. The path processing unit is used to adapt to file path formats of different operating systems, including compatibility processing for long paths and automatic escaping of path separators, to ensure that the script can be executed normally in different system environments.

10. An intelligent video automatic generation system based on random chapter combination, applied to the intelligent video automatic generation method based on random chapter combination as described in any one of claims 1-9, characterized in that, include: User interface module, communication middleware layer, core algorithm module, professional rendering module; The user interface module is used to provide a visual operation interface, receive chapter parameter configuration input by the user, which corresponds to the chapter parameters in step 1 of the method, and simultaneously display the number of video combinations calculated by the core algorithm module in real time, which corresponds to the number of video combinations in step 2 of the method. The communication middleware layer is used to establish a communication link between the user interface module and the core algorithm module, realize the bidirectional transmission of chapter parameters, calculation instructions and result data, and manage file system access permissions to ensure the read and write security of materials and generated files. The core algorithm module is used to execute the core logic of steps two to five of the method. Specifically, it includes calling the corresponding combination number calculation algorithm according to the chapter mode, generating a timeline containing audio-visual alignment parameters and video effect parameters, exporting an open timeline exchange format standard file, and generating an automated script for professional video editing software. The professional rendering module is used to interface with mainstream professional video editing software, receive automated scripts generated by the core algorithm module, and execute the material import, timeline construction, and batch rendering commands in the scripts. This process corresponds to step six of the method, and finally completes the automated export of the video.

Images