Event log-based live stream reconstruction method, computing device, and storage medium

Abstract

This invention discloses a live stream correction and reconstruction method and a filtering and reconstruction method based on event logs, aiming to solve the problems of secondary modification and content traceability difficulties caused by the high coupling of content data in traditional live streaming and recording processes. The method adopts a data stream separation architecture, capturing and separating the original video frame stream and accompanying metadata in real time. In the content processing stage, the method performs semantic-level content decapsulation and reorganization of the original data stream based on highly structured metadata instructions. This solution achieves modularization of the content processing workflow, enhances content traceability, modification flexibility, and applicability, supports multi-dimensional deconstruction and refined secondary editing of original materials, thereby supporting live stream playback traceability, recording process reorganization, and content asset reuse, significantly reducing the technical threshold and time cost of later maintenance.

Description

Technical Field

[0001] The embodiments of this application relate to the field of video live streaming technology, and in particular to a live stream reconstruction method, computing device, and storage medium based on event logs. Background Technology

[0002] In the field of video live streaming technology, especially in systems involving real-time streaming media processing, live content generation, or interactive audiovisual content recording, existing solutions generally face a core deficiency: the lack of ability to generate and save structured metadata for "post-reconstruction" during the recording process.

[0003] Specifically, current mainstream consumer and professional systems typically only output the final, rendered composite video file when recording composite video streams containing multiple layers of overlay content (such as real-time comments, graphic icons, interactive bullet comments, dynamic effects, game UI, etc.). When the final composite streaming media is lost during transmission, corrupted in storage, or requires post-production modifications (such as correcting errors in the overlays, adjusting effect positions, or complying with new content guidelines), existing technologies cannot provide efficient correction solutions. Content creators or engineers have no actionable intermediate data for targeted repair. Their only option is a completely manual, re-editing process starting from the original footage. This not only requires reintroducing all overlay elements and ensuring absolute precision in time synchronization and spatial positioning, but the entire process is extremely time-consuming, labor-intensive, and highly dependent on manual operation, failing to guarantee consistency with the original output and severely impacting content production efficiency and post-processing flexibility. Summary of the Invention

[0004] To address the technical challenges of secondary modifications and content traceability caused by the high coupling of content data in the aforementioned live streaming and recording processes, this invention employs a dual-output pipeline architecture for live streams that simultaneously generates unmodified original video data and enhanced live streaming data from a single video capture source. This architecture includes an event log synchronization engine, enabling frame-accurate enhanced output and post-construction from the original video data in a corrective or filtered manner.

[0005] A first aspect of this application provides a live stream correction and reconstruction method based on event logs, wherein the event logs are logs of enhancement events added to the original video during the live stream according to preset rules, and the method includes: Obtain the raw video data from the live stream; Based on the input correction instructions, a correction process is performed on the copy of the event log to obtain the target event log; Based on the original video data and the target event log, the reconstructed video data is rendered.

[0006] Preferably, the event log includes at least a unique identifier for the enhanced event, start and end timestamps, and enhanced content parameters.

[0007] Furthermore, the enhanced content parameters include at least the type of the enhanced event, the content parameters, and the screen position.

[0008] The event log is generated synchronously with the original video data and the live stream video through a data synchronization module during the live broadcast. The data synchronization module uses a shared frame reference with assigned frame sequence numbers and absolute timestamps as a reference to collect and aggregate the rendering parameter set from the live stream video pipeline in real time as the event log.

[0009] A second aspect of this application provides a live stream filtering and reconstruction method based on event logs, wherein the event logs are logs of enhanced events added to the original video during the live stream according to preset rules, and the method includes: Obtain the raw video data from the live stream; Obtain a filtering instruction, wherein the filtering instruction is an instruction to hide one or more of the enhanced events; Based on the filtering instructions, the original video data, and the event log, the reconstructed video data is rendered.

[0010] Preferably, the event log includes at least a unique identifier for the enhanced event, start and end timestamps, and enhanced content parameters.

[0011] Furthermore, the enhanced content parameters include the type of the enhanced event, content parameters, and screen position.

[0012] The event log is generated synchronously with the original video data and the live stream video through a data synchronization module during the live broadcast. The data synchronization module uses a shared frame reference with assigned frame sequence numbers and absolute timestamps as a reference to collect and aggregate the rendering parameter set from the live stream video pipeline in real time as the event log.

[0013] A third aspect of this application discloses a computing device including a memory and a processor, wherein the memory stores a computer program, characterized in that the processor executes the computer program to implement the method described in the first or second aspect of this application.

[0014] A fourth aspect of this application discloses a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the method described in the first or second aspect of this application.

[0015] The event log-based live stream correction and reconstruction method described in this application enables precise retrospective and targeted modification of live stream content without re-collecting the original materials. The core advantage of this method lies in independently saving the description information of enhanced events in a structured log format, rather than embedding it in the final video frames, thus constructing a separable, editable, and reusable content production system. When issues such as typos in subtitles, misaligned effects, inappropriate bullet comments, or misaligned interactive elements occur during the live stream, operators do not need to manually fix them frame by frame. They only need to locate the corresponding event log entry, adjust relevant parameters or time attributes through correction commands, or hide the corresponding event through filtering commands. The system can then automatically re-render based on the original video data, generating reconstructed video data that meets the requirements. This architecture significantly reduces the technical threshold and time cost of post-maintenance, while ensuring a high degree of visual consistency between the modified output and the original live stream, avoiding synchronization errors or style deviations that may be introduced by manual re-production. Attached Figure Description

[0016] Figure 1 A data flow diagram of a live streaming process based on an embodiment of this application is shown;

[0017] Figure 2 A schematic diagram of a terminal device or server suitable for implementing embodiments of this application is shown. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0019] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three possibilities exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0020] The live stream reconstruction method provided in this application can run on various computing devices, including laptops, smartphones, tablets or other intelligent devices with computing capabilities, such as smart cameras, wearable devices, etc., or it can run in the cloud and provide live stream reconstruction services through network communication.

[0021] The live stream reconstruction method provided in this application is based on the original video stream data and synchronized video logs in the live stream. The video logs refer to the logs of enhancement events that are synchronously identified and added to the original video during the live stream.

[0022] In one embodiment of this application, the following four operations are performed simultaneously during live video streaming.

[0023] 1. Preserve the original video data of the live stream. The original video data is the original video frame sequence data without any enhancement events. This data does not contain any text, graphics, special effects or other elements added later, and completely preserves the original image information captured by the image sensor in the live stream scene.

[0024] Second, the original video data is analyzed and processed to obtain enhancement events to be added based on preset rules. These enhancement events are then overlaid on the original video data to create a real-time live stream. Specifically, enhancement events can be automatically triggered by the system, such as program logos or time watermarks added at specific times according to preset rules; they can also be generated by user interaction, such as bullet comments sent by viewers or interactive effects triggered by the broadcaster; or they can be obtained through analysis of the original video information, such as highlight trajectories, sports scores, replays of highlights or picture-in-picture sequences, or digital commentary.

[0025] The preset rules are a set of predefined, configurable automated triggering logics. They specify the response actions the system should automatically execute when a specific event occurs. These rules are the core mechanism for enhancing the system's intelligence, allowing users to decouple and map triggering conditions with execution results based on different business scenarios and specific business objectives. In different embodiments, the preset rules can be time-based, such as automatically adding a marker 5 minutes after the start of the live stream; they can also be content-based, such as automatically displaying a personal information card in the lower right corner of the screen when a specific person appears in the original video, or triggering a related image and text pop-up when a preset keyword is detected in the screen; or they can be user interaction-based, such as automatically displaying a dynamic "High Energy Comments" prompt at the top of the screen when the number of comments sent by viewers reaches a preset threshold (e.g., 10 per second). These preset rules can be flexibly defined and modified through configuration files, visual interfaces, or programming interfaces to adapt to the personalized needs of different live streaming scenarios such as sports, games, e-commerce, and education. For example, in sports live streaming scenarios, preset rules can be set to display a text bar containing information such as the event type, time of occurrence, and players involved at the bottom of the screen when key match events such as goals, fouls, and substitutions are detected. During match breaks, enhanced events such as real-time scores, possession rate, and number of shots will be automatically displayed. The style and display position of the enhanced events can be adjusted according to the type of event (such as football, basketball, and tennis), for example, highlighting the corner kick area in football live streams and displaying trajectory effects for three-point shots in basketball live streams. In e-commerce live streaming scenarios, preset rules can be set to automatically display enhanced content parameters such as the name, price, and stock quantity of a product on the left side of the screen when the host mentions it, and trigger an enhanced event with magnification effects when the host displays product details. In game live streaming scenarios, preset rules can be set to display enhanced events such as kill information and score changes in the center of the screen when a game character completes a specific operation, such as a kill, and automatically switch different UI layouts according to the game progress.

[0026] Third, maintain a log of the enhanced events, that is, organize the aforementioned enhanced events into a structured event log in chronological order of occurrence. Each log entry should record at least the unique identifier of the enhanced event, the start and end timestamps, and detailed enhanced content parameters, such as event type (subtitles, bullet comments, special effects, etc.), specific content information (text content, image paths, special effect parameters, etc.), and the precise location coordinates in the video frame.

[0027] Fourth, based on the original video data and real-time generated event logs, a live stream video for real-time viewing is synchronously rendered. This live stream video is the final presentation effect after overlaying the original video with various enhanced events. In this way, the generation and storage of three key data sets—original video data, event logs, and live stream video—are completed synchronously during the live broadcast, forming the data foundation for live stream reconstruction. In practical applications, this synchronous generation mechanism ensures strict timeline alignment between the event logs and the original video data, providing frame-level precise synchronization for subsequent reconstruction operations. An example is provided below to illustrate this.

[0028] For example, when a subtitle "Welcome new viewers to the live stream" needs to be added during a live stream from 00:10:23 to 00:10:45, the system will create a record in the event log containing a unique ID for the subtitle, such as "Event_20231001_001", start and end timestamps "00:10:23.000 to 00:10:45.000", event type classification as "subtitle", and content parameters including the text "Welcome new viewers to the live stream", font "Microsoft YaHei", font size "24", color "#FFFFFF", and screen position (e.g., X:500, Y:800). Simultaneously, the original video data will record all original screen information frame-by-frame from 00:10:23 to 00:10:45, and the live stream video is formed by overlaying this subtitle information onto the original screen in real time according to the parameters in the log. This data generation method tightly links the original video data, event log, and live stream video in the time dimension.

[0029] In one possible implementation, the event log is generated synchronously with the original video data and the live stream video during the live broadcast via a data synchronization module. This data synchronization module uses a shared frame reference with assigned frame sequence numbers and absolute timestamps as a reference, and synchronously collects and aggregates the rendering parameter set from the live stream video pipeline as the event log in real time. Specifically, this live broadcast process can be implemented using a dual-pipeline video architecture with a synchronization module. The dual-pipeline video processing architecture is a highly optimized real-time video stream generation system. Its core function is to achieve parallel, asynchronous, and high-precision processing of the same original video source, and synchronously record event metadata for all enhanced content. This system mainly consists of a frame acquisition module, a first video pipeline, a second video pipeline, and a data synchronization module.

[0030] First, the frame acquisition module, acting as the system's front-end interface, directly interacts with the sensor of the shooting device to acquire raw frames from the camera sensor, thus obtaining the raw video stream. Its key lies in its data memory management mechanism, employing WCS (Write Copy On Write) and DMA (Direct Memory Access) level memory operations to process each captured video frame, generating two logically independent frame buffer references for each frame: a first frame reference and a second frame reference. Notably, this process does not physically copy pixel data, generating only minimal metadata overhead. This ensures that within extremely low latency (less than 1 millisecond), the two subsequent processing pipelines can access byte-level consistent and time-synchronized memory addresses pointing to the same raw image snapshot, guaranteeing minimal physical redundancy and extremely high processing efficiency.

[0031] Subsequently, the system activates two video processing paths in parallel. The first video pipeline, based on the first frame buffer reference, is responsible for generating the raw video stream. The core function of this pipeline is to use a hardware video encoder to encode the raw video frames losslessly or with minimal loss. Its output goal is to generate a completely clean reference video stream, untouched by any post-rendering intervention, as a baseline record of the original events. Conversely, the second video pipeline, based on the second frame buffer reference, is responsible for generating the enhanced rendered video stream. This pipeline inputs video frames into a pre-defined, multi-layered rendering engine. This rendering engine, as the core functional unit, maintains an ordered rendering layer stack, performs precise compositing through the image processor's hybrid pipeline, and calculates rendering parameters including elements such as watermarks, status indicators, and scoreboards. Finally, these rendering parameters, along with the raw video frames, are encoded by the hardware video encoder to form the final output video stream containing real-time enhancement information.

[0032] The data synchronization module, acting as a scheduling and data aggregation unit, is tasked with accurately recording events across pipelines and frames, and creating an event log. Once the frame acquisition module assigns a shared frame reference with a sequence number and absolute timestamp to a particular frame, the data synchronization module intervenes. Using this shared frame reference as a timeline reference, it receives rendering parameters generated by the second video pipeline. It structurally records all metadata related to rendering events, including unique identifiers, timestamps or time ranges, activated compositing layers, enhancement content types, screen pixel coordinates, complete JSON-formatted content parameter sets, trigger sources, and animation timeline parameters, generating a unified rendering parameter integration file. This integration file is essentially a high-precision, frame-based event log. By providing this extremely fine-grained, machine-readable event recording, this module significantly improves the auditability and traceability of the video system, ensuring that every enhancement decision and parameter change during video content generation can be permanently and accurately captured and stored.

[0033] In one possible implementation, based on the aforementioned dual-pipeline video architecture, taking ball games as an example, the specific process and data flow during live streaming are detailed below.

[0034] refer to Figure 1 The data stream originates from the raw data input from the camera sensor. The camera sensor captures the raw light signal of the target scene in real time and generates a raw video stream. Upon receiving this video stream, the video processing system first uses the DMA pipeline allocator module to perform memory management operations on each video frame. A key function of this module is to generate and maintain a shared frame reference. This shared frame reference is not only a memory address but also a lightweight descriptor that must contain the physical address information of the video frame, a unique frame sequence number, and a precise absolute timestamp. Through this mechanism, the system ensures that all subsequent processing pipelines can perform multi-dimensional parallel access to the same video frame with extremely low memory overhead and extremely short latency, thereby achieving strictly synchronized raw image snapshots in both time and space.

[0035] The data stream then enters the first processing path, the first video pipeline, which processes the raw video data. This first video pipeline acquires raw video frame data based on a shared frame reference and inputs it into a high-level hardware video encoder, such as H.265. The sole objective of this pipeline is to generate a high-fidelity raw reference video stream free from any graphic overlays, text, or status information interference. This raw video stream is written to the device's persistent storage unit using a standard container format, serving as a pure and immutable archive record of the original scene.

[0036] When the first processing path is activated, the system simultaneously activates the second processing path, namely the second video pipeline for processing the live stream. This path relies on a series of intelligent modules: First, the AI ​​highlight detection module analyzes the raw video frames, identifies key events, and triggers system state changes. Once triggered, the system obtains real-time scoreboard data and data binding information from an external data source. This data guides the rendering process, and the rendering results are processed by the enhanced compositing GPU module. This GPU module is a multi-layer rendering system that receives the raw video frames and combines them with rendering parameters provided by upper layers (such as layer 0: raw frame; layer 1: watermark; layer 2: scoreboard; layer 3: identity card), performing compositing calculations through the GPU's hybrid pipeline to generate composite frames with real-time enhanced visual effects. Finally, this composite frame data is sent to a second hardware video encoder, encoded, and output as a real-time live stream for users to receive and watch in real time.

[0037] While both processing paths are activated simultaneously, the system also activates the data synchronization module. Throughout the process, all rendering parameter calculations, including which enhancements are invoked, where they are displayed, which data bindings are used, and the log structure of the content parameters, are captured by the data synchronization module. As the data stream's convergence point, the data synchronization module uses the original shared frame reference as the timeline anchor point and structurally integrates all rendering parameters calculated by the second video pipeline with the corresponding trigger information (including the trigger source and the complete set of enhancement parameters). This process generates an enhancement content parameter log file, which is an event log strictly synchronized with the video frame timeline. This log file records all metadata for each enhancement event, enabling the system to accurately map video frame sequence numbers / timestamps to enhancement event parameters one-to-one, thus achieving synchronous generation and associated storage of the live stream and event logs.

[0038] In one possible implementation, the enhanced content parameters are multi-dimensional, aiming to comprehensively describe the identity, location, data content, triggering mechanism, and dynamic performance of the enhanced content. Specifically, these parameters include defining the category of the enhanced content type, such as bullet comments, logos, identity cards, scoreboards, etc., its specific layer in the composite layer, and its precise pixel bounding box coordinates at the output resolution. The core content parameter set is the foundation for reconstructing the enhanced content; it contains information such as template ID, data binding, animation status, and resource references, ensuring the integrity and reusability of the enhanced content. Furthermore, the enhanced content parameters can also cover the description of the triggering mechanism and dynamic behavior. The trigger source and related trigger-specific data jointly define the specific reason for the appearance of the enhanced content, such as due to a global scene event, AI highlight recognition, timed scheduling, or manual request. Finally, in a preferred embodiment, the dynamic performance of the enhanced content can be described in detail through animation timeline parameters, including entrance and exit transition effects, duration, and interpolation parameters per frame, thus completely defining the entire lifecycle of the enhanced content from its appearance to its disappearance. The following example of a specific sports scene illustrates the possible composition of the enhanced content parameters.

[0039] In a live football broadcast, when the system triggers a "goal" enhancement event through AI highlight recognition, the enhancement content parameters can be specifically configured as follows: the event type is categorized as "highlight event alert," the layer level is set to "1," and the pixel bounding box coordinates are defined as (X: 100, Y: 50, width: 300, height: 80) to reserve a specific area for display in the upper left corner of the screen. The core content parameter set includes the template ID "Goal_Alert_Template_003" (calling the preset goal alert template), the data binding fields are associated with "Goal scorer's name (Kobe)," "Goal time (65')," and "Current score (2-1)" in the real-time event database, the animation status is set to "inactive," and the resource reference points to the team logo image path " / ** / ** / teamA.png." The trigger mechanism description identifies the trigger source as "AI_Event_Detector". The trigger's specific data includes the original video frame timestamp "01:05:23.456" at the moment of the goal, the goal area coordinates (X: 400-500, Y: 250-350), and the motion trajectory vector (Vx: 15, Vy: -8). The animation timeline parameters are defined in detail: the entrance transition effect is "slide in from the right," lasting 0.5 seconds; the duration is set to 5 seconds; the exit transition effect is "fade away," lasting 0.3 seconds; the interpolation parameters for each frame use the "easeOutQuart" easing function to ensure a natural and smooth animation. Through this multi-dimensional parameter configuration, the system can accurately reproduce the complete presentation of the goal event in the live broadcast, including when it occurs, its location and display style, the dynamically updated match data included, and how to perform smooth entrance and exit animations.

[0040] Based on the original video data and event logs mentioned above, the live stream can be reconstructed by rebuilding or filtering the event logs to form an updated live replay video.

[0041] In one embodiment of the present invention, a target event log is obtained by correcting the event log, and the live stream is reconstructed based on the original video data and the target event log. Specifically, firstly, the original video data in the live stream is acquired. This original video data is a sequence of original frame images during the live stream without the overlay of enhanced events, fully preserving the original visual information of the live scene. Secondly, a correction instruction input by the user is acquired. This correction instruction indicates the adjustment operation to be performed on a specific enhanced event in the event log, such as modifying the text content of the subtitles, modifying the scoreboard, adjusting the display duration of special effects, and correcting the position coordinates of graphic icons. Next, based on the correction instruction, a copy of the event log is corrected to obtain the target event log. During the correction process, the system locates the corresponding log entry in the event log copy according to the unique identifier of the enhanced event specified in the correction instruction, and then modifies the start and end timestamps or enhanced content parameters of the entry according to the requirements of the correction instruction. For example, if the correction instruction is to change the subtitle content of event ID "Event_20231001_001" from "Welcome new viewers to the live stream" to "Welcome all new and old friends to the live stream," the system will find the entry corresponding to that ID in the event log copy and update the "Text Content" field in its "Content Parameters" to the corrected text. If the correction instruction involves adjusting the start and end times of an event, such as extending the end time of a special effect from 00:10:45 to 00:10:55, the system will modify the "Start and End Timestamp" field of that event log entry accordingly. After completing the correction of the event log copy, the resulting target event log contains all the corrected enhanced event information. Finally, based on the original video data and the target event log, the reconstructed video data is rendered. During the rendering process, the system reads the original video data frame by frame and, according to the start and end timestamps and enhanced content parameters of each enhanced event in the target event log, precisely overlays the corrected enhanced events onto the corresponding video frames. For example, for the modified subtitle event mentioned above, the system will overlay the updated text "Welcome all new and old friends to the live stream" onto the original video frame within the time range of 00:10:23.000 to 00:10:45.000, according to the specified font, font size, color, and position coordinates, thereby generating corrected reconstructed video data. This reconstructed video data not only preserves the picture quality of the original video but also corrects the problems existing in the enhanced event, realizing accurate backtracking and targeted repair of the live stream. In this embodiment, the correction instruction also includes the instruction to hide the enhanced event.

[0042] In another embodiment of the invention, if no modification to the enhanced events is required, but only some enhanced events need to be hidden, filtering instructions for the enhanced events can be received. During re-rendering, the live stream is reconstructed based on the filtering instructions, the original video data, and the original event logs. In this embodiment, since there is no need to regenerate the target event logs, the efficiency of rendering and reconstructing the video data can be improved. Specifically, firstly, the original video data in the live stream is obtained. This data is the original frame sequence of the live stream without any enhanced events superimposed, completely recording the original visual information of the live scene. Secondly, the filtering instructions input by the user are obtained. These instructions explicitly indicate one or more enhanced events that need to be hidden, such as hiding bullet comments during a specific time period, removing an incorrect special effect icon, or deleting a non-compliant subtitle. The filtering instructions can be input through the user interface, specifying a unique identifier for the enhanced events to be hidden, or batch selecting events to be hidden based on conditions such as event type and time range. Next, during the rendering and reconstruction of the video data, the original video data, event logs, and filtering instructions are read synchronously. When an enhanced event in the event log read by the system also has a corresponding filtering instruction, the system automatically ignores it and does not render it into the reconstructed video.

[0043] For example, if the filtering instruction is to hide the bullet comment event with the event ID "Event_20231001_002", then during the rendering process, when processing video frames within the start and end timestamp range corresponding to this event, the system will skip the overlay step of the bullet comment event and only retain the original video frame. If the filtering instruction is based on a time range, such as hiding all interactive effect events between 00:15:00 and 00:16:00, then the system will traverse all entries in the event log within this time range and of type "interactive effect", excluding all enhanced events of these entries during rendering. This reconstruction method based on filtering instructions does not require modification of the original event log, but only dynamically determines whether to overlay specific enhanced events during the rendering stage, thereby quickly generating reconstructed video data that does not contain the specified enhanced events. It is particularly suitable for handling illegal content discovered after a live broadcast, sensitive information that needs to be temporarily hidden, or enhanced elements that do not need to be permanently deleted but need to be blocked in a specific version, making it convenient for maintaining live broadcast playback videos.

[0044] In another embodiment of the present invention, a target event log can be obtained by extending the event log, and the live stream can be reconstructed based on the original video data and the target event log, thereby supplementing or upgrading the content of the original live stream. Specifically, firstly, the original video data in the live stream is acquired, which serves as the basic frame and provides complete visual information of the live scene. Secondly, an extension instruction input by the user is acquired, which indicates the enhancement events to be added to the event log, such as adding new narration subtitles during a specific time period, inserting dynamic advertisements from sponsors, overlaying real-time statistical data charts, or supplementing historical review segments with identifiers. The extension instruction must include detailed information about the new enhancement event, such as the event type, target timestamp or time range, enhancement content parameters (such as text content, image resources, location coordinates, animation effects, etc.), and trigger source. Next, based on the extension instruction, an extension process is performed on a copy of the event log to obtain the target event log. During the extension process, the system generates a unique identifier for the new enhancement event and inserts a new log entry at the corresponding timeline position in the event log copy according to the content of the extension instruction. For example, if the extension instruction is to add a statistical caption about "total career goals" between 00:20:15.000 and 00:20:30.000, the system will create a new event entry in the event log copy. Its unique identifier might be "Event_20231001_New001", the event type would be categorized as "statistics", the start and end timestamps would be set to 00:20:15.000 to 00:20:30.000, and the content parameters would include the text "total career goals: 128", font, color, and location coordinates (e.g., X: 500, Y: 400), with the trigger source labeled "Manual_Extension". If the extension instruction involves multiple new events or complex dynamic sequences, the system will insert the corresponding log entries sequentially according to time order to ensure the correct temporal logic between events. After extending the event log copy, the resulting target event log will contain all the newly added enhanced event information. Finally, based on the original video data and the target event log, the reconstructed video data is rendered. During rendering, the system reads the original video data frame by frame and precisely overlays all enhancement events onto the corresponding video frames based on the start and end timestamps and enhancement content parameters of all original and newly added enhancement events in the target event log. For newly added statistical caption events, the system will overlay "Career Total Goals: 128 Goals" onto the original video frame according to the specified style and position within the time range of 00:20:15.000 to 00:20:30.000.This method of reconstructing live streams by extending event logs can flexibly add new information elements, advertising content, or interactive functions to live replay videos without changing the original video data, thereby enhancing the added value and watchability of the videos. It is especially suitable for producing event highlights, creating secondary content, or meeting the customized needs of different platforms and audiences.

[0045] Figure 2 A schematic diagram of a terminal device or server suitable for implementing embodiments of this application is shown.

[0046] like Figure 2 As shown, the terminal device or server includes a central processing unit (CPU) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from storage section 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the terminal device or server. The CPU 501, ROM 502, and RAM 503 are interconnected via bus 504. An input / output (I / O) interface 505 is also connected to bus 504.

[0047] The following components are connected to I / O interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to I / O interface 505 as needed. A removable medium 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 510 as needed so that computer programs read from it can be installed into storage section 508 as needed.

[0048] Specifically, according to embodiments of this application, the above method flow steps can be implemented as a computer software program. For example, embodiments of this application include a computer program product comprising a computer program carried on a machine-readable medium, the computer program containing program code for performing the methods shown in the flowchart. In such embodiments, the computer program can be downloaded and installed from a network via communication section 509, and / or installed from removable medium 511. When the computer program is executed by central processing unit (CPU) 501, it performs the functions defined in the system of this application.

[0049] It should be noted that the computer-readable medium shown in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media can also be any computer-readable medium other than computer-readable storage media, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, or any suitable combination thereof.

[0050] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation that may be implemented in systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0051] The units or modules described in the embodiments of this application can be implemented in software or hardware. The described units or modules can also be located in a processor. The names of these units or modules do not, in certain circumstances, constitute a limitation on the unit or module itself.

[0052] In another aspect, this application also provides a computer-readable storage medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium stores one or more programs that are used by one or more processors to execute the methods described in this application.

[0053] In another aspect, embodiments of this application also provide a computer program product that, when executed by a processor, implements the methods of any of the above embodiments.

[0054] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the foregoing application concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions claimed in this application.

Claims

1. A live stream correction and reconstruction method based on event logs, characterized in that, The event log is a log of enhanced events added to the original video during live streaming according to preset rules. The method includes: Obtain the raw video data from the live stream; Based on the input correction instructions, a correction process is performed on the copy of the event log to obtain the target event log; Based on the original video data and the target event log, the reconstructed video data is rendered.

2. The method as described in claim 1, characterized in that, The event log includes at least the unique identifier of the enhanced event, start and end timestamps, and enhanced content parameters.

3. The method as described in claim 1, characterized in that, The event log is generated synchronously with the original video data and the live stream video through a data synchronization module during the live broadcast. The data synchronization module uses a shared frame reference with assigned frame sequence numbers and absolute timestamps as a reference to collect and aggregate the rendering parameter set from the live stream video pipeline in real time as the event log.

4. The method as described in claim 2, characterized in that, The enhanced content parameters include at least the type of the enhanced event, content parameters, and screen position.

5. A live stream filtering and reconstruction method based on event logs, characterized in that, The event log is a log of enhanced events added to the original video during live streaming according to preset rules. The method includes: Obtain the raw video data from the live stream; Obtain a filtering instruction, wherein the filtering instruction is an instruction to hide one or more of the enhanced events; Based on the filtering instructions, the original video data, and the event log, the reconstructed video data is rendered.

6. The method as described in claim 5, characterized in that, The event log includes at least the unique identifier of the enhanced event, start and end timestamps, and enhanced content parameters.

7. The method as described in claim 5, characterized in that, The event log is generated synchronously with the original video data and the live stream video through a data synchronization module during the live broadcast. The data synchronization module uses a shared frame reference with assigned frame sequence numbers and absolute timestamps as a reference to collect and aggregate the rendering parameter set from the live stream video pipeline in real time as the event log.

8. The method as described in claim 6, characterized in that, The enhanced content parameters include at least the type of the enhanced event, content parameters, and screen position.

9. A computing device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.

Images