Video clip automatic generation method and system based on match score data driving

Abstract

The invention discloses a video clip automatic generation method and system based on match score data driving, and relates to the technical field of video processing and sports matches. The method comprises the steps that score data generated by an intelligent scoring terminal in the table tennis competition process are collected in real time, wherein the score data comprise the occurrence moment of each score and the score state of the moment; acquiring video stream data in time synchronization with the score data; automatically identifying a target moment from the score data according to a preset editing rule; and intercepting a video clip in a corresponding time range from the video stream data according to the target moment. The preset editing rules comprise various types such as a whole game, playback of each score, winning score collection, key score fragments, reverse game fragments and dispute playback fragments, dispute playback is triggered and checked through mobile terminal equipment, and players can select to check the video clips of the current score, the previous score or the first two scores according to requirements.

Description

Technical Field

[0001] This invention relates to the fields of video processing and sports event technology, specifically to a method and system for automatically generating video clips based on sports score data. Background Technology

[0002] With the widespread adoption of short videos and social media, the demand for video content related to sporting events is increasing daily. For score-driven sports like table tennis, highlight moments (such as crucial points, winning points, and comebacks) have extremely high entertainment and dissemination value. However, current technologies for editing sports event videos mainly suffer from the following problems: I. Inefficiency of Manual Editing: Traditional video editing relies on manual viewing of the entire recording, marking highlights, and extracting segments. A table tennis match can last for several hours, making manual editing time-consuming, labor-intensive, and unable to cover all the exciting moments. For amateur events, the high cost of editing hinders the widespread adoption of video services.

[0003] II. Limitations of existing automatic editing technology: 1. The video editing method based on the positional relationship parameters of the ball in the game relies on image recognition to analyze the ball's trajectory, which is computationally complex and easily affected by light occlusion. 2. Determine the target video slice from multiple video streams based on time points. This only addresses time point location and does not cover how to automatically identify highlights. 3. Artificial intelligence-based football match editing methods rely on complex AI models such as object detection networks and temporal semantic networks, requiring a large amount of training data; 4. The AI ​​real-time football match editing system relies on face engine, OCR engine, and 3D engine, which is technically complex and expensive; 5. Editing based on the start time of the game as seen in the scoreboard image relies on professional scoreboard information in the video and is not suitable for amateur events without professional scoreboards. 6. Video editing for sporting events based on tracking gimbal data relies on gimbal yaw angle data to select exciting clips, requiring professional tracking and shooting equipment.

[0004] III. Common technological gaps in existing technologies: 1. Both rely on video content analysis: Whether it is AI visual recognition or image detection, existing technologies all require analysis of video content, which is computationally complex, costly, and greatly affected by factors such as lighting and occlusion.

[0005] 2. Inability to fully utilize match data: Existing technology fails to fully utilize the high-value information source of score data already existing during the match.

[0006] 3. Not suitable for amateur events: Professional-grade AI systems are too expensive for amateur events to afford; while systems that rely on professional scoreboards are not suitable for amateur events without professional scoreboards. Summary of the Invention

[0007] To overcome the shortcomings of the prior art, the present invention provides the following technical solution: a method for automatically generating video clips based on sports score data, characterized by comprising: S1, real-time acquisition of score data generated by a smart scoring terminal during a sports event, wherein the score data includes the time of occurrence of each point and the score status at that time; S2, acquisition of video stream data synchronized with the score data; S3, identification of a target time from the score data according to preset editing rules; S4, extraction of video clips within the corresponding time range from the video stream data according to the target time, generating a target video clip; S5, tagging and storage, setting tags for the generated target video clips to identify their corresponding event types, and pushing them to the cloud or local storage for users to access via mobile devices.

[0008] Preferably, the preset editing rules include at least one of the following: The entire match rules are as follows: the first preset time before both players take their positions and trigger their actions is taken is taken as the start time, and the second preset time after the end time of the match is taken as the end time. The entire match video segment is extracted. Playback rules for each minute: Centered on the moment when each minute occurs, extract video segments of preset duration before and after that moment to generate an independent playback segment for each minute; the preset duration can be flexibly configured according to actual needs, including but not limited to 1 second / 2 seconds / 3 seconds / 5 seconds / 10 seconds; Winning Point Highlights Rules: Identify all winning points that end a game, extract video clips of preset duration before and after each winning point, and generate winning point highlights; Key score rule: When the score reaches the preset key score value (such as 10:10, 12:12, etc.), extract video segments of preset duration before and after each subsequent point to generate key score segments; Comeback Rules: Identify the score change range from falling behind to taking the lead, extract a video segment from the start of the range to the end of the range with a preset duration, and generate a comeback segment; Dispute replay rules: In response to a dispute replay request triggered by a user via a mobile device, a video segment with the corresponding score is extracted from the video stream data according to the replay range selected by the user, and a dispute replay segment is generated; the replay range includes the current score, the previous score, or the previous two scores.

[0009] Preferably, the time synchronization of the score data and the video stream data is achieved in the following ways: synchronizing the time of the intelligent scoring terminal and the time of the video acquisition device to the same time reference server; or the intelligent scoring terminal sending a synchronization signal containing a timestamp to the video acquisition device at each point, and the video acquisition device establishing an index relationship between the timestamp and the video frame.

[0010] Preferably, it further includes: setting a tag for the generated target video clip, the tag being used to identify the event type corresponding to the video clip, the event type including: the whole game, every point, winning point, key point, comeback, and controversial replay; It also includes: pushing the generated target video clips to cloud storage or local storage; providing an interface for users to access via mobile devices, supporting users to watch, download or share the video clips online; The intelligent scoring terminal includes trigger buttons located on both sides of the ping-pong table, which are used to respond to the player's hitting action to update the score and record the moment each point occurs.

[0011] Preferably, the action of both players taking their positions includes clicking a button, double-clicking a button, pressing both buttons simultaneously, or confirming their positions via a mobile device; the end time of the match is determined as follows: after the last point of a game is detected, and no new score data is available within a preset time period, the moment when the last point occurs is taken as the end time of the match; the dispute replay request is triggered by an application or mini-program on the mobile device, and the user can select to view the video clip of the current point, the previous point, or the previous two points by clicking the replay option on the interface.

[0012] Preferably, it includes: at least one intelligent scoring terminal for real-time collection of score data during a sports event, the score data including the time of each point and the score status at that time; at least one video acquisition device for collecting video stream data synchronized with the score data; and at least one backend server, communicatively connected to the intelligent scoring terminal and the video acquisition device, for receiving the score data and the video stream data. The backend server includes: a data receiving module for receiving score data and video stream data; an editing rule module for storing preset editing rules; a target time identification module for identifying a target time from the score data according to the preset editing rules; a video extraction module for extracting video segments within a corresponding time range from the video stream data according to the target time; a video storage module for storing the generated target video segments; and an access interface module for providing video segment access services to mobile devices.

[0013] Preferably, it also includes a mobile device, which runs an application or mini-program for receiving a dispute replay request triggered by a user, displaying a replay option interface, and playing a dispute replay segment obtained from a backend server.

[0014] Preferably, the resolution of the video acquisition device is sufficient to clearly identify the range of the ball table, without limiting the specific resolution value.

[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for automatically generating video clips based on match score data.

[0016] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement any of the above-described methods for automatically generating video clips based on match score data.

[0017] Compared with the prior art, the present invention has the following advantages: 1. No AI visual analysis is required, significantly reducing costs. This invention does not rely on complex image recognition, object detection, AI models, or other technologies. It can automatically edit videos using only existing score data, greatly reducing system complexity and costs. The hardware cost of a single camera is controllable, and storage costs can be effectively controlled through intelligent recording strategies.

[0018] 2. High editing precision and real-time performance: Because the timestamp accuracy of the score data is down to the millisecond level, the start and end points of video clipping can be precisely controlled, avoiding errors inherent in manual editing. Furthermore, editing can be completed instantly after the match, achieving "immediate output after the match."

[0019] 3. Flexible and diverse editing rules to meet the needs of multiple scenarios. By preset different editing rules, various types of video content can be automatically generated, such as the whole game, replays of each point, highlights of winning points, key point segments, comeback segments, and controversial replay segments, to meet the needs of different roles such as players, audiences, and sponsors.

[0020] 4. The dispute replay design is user-friendly and suited to the realities of amateur competitions. Dispute replays are triggered and viewed via mobile devices, eliminating the need for additional dedicated display equipment and lowering the barrier to entry for venues. Players can flexibly choose to view video clips of the current score, the previous score, or the previous two points, satisfying on-site verification needs while avoiding conflicts with other functions of the scoring terminal.

[0021] 5. Provides professional-grade video service capabilities for amateur events. This invention has low deployment costs, requires no professional camera team or AI server, and can provide professional-grade video service capabilities for amateur table tennis events, including automatic generation of highlights, real-time replay of controversial balls, and online live replay.

[0022] 6. Maximizing the value of data assets: This invention "reutilizes" score data that was originally only used for scoring, making it the core driving force for video editing, thereby maximizing the value of data assets.

[0023] 7. This invention establishes a clear technological distinction from existing technologies, which all rely on video content analysis, while this invention relies on score data-driven approaches—the technical problems they solve, the technical methods they employ, and the technical effects they achieve are completely different. This invention fills the technological gap of "using score data to drive video editing." Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating the overall process of the method of the present invention.

[0025] Figure 2 This is a schematic diagram of the system architecture of the present invention.

[0026] Figure 3 This is a flowchart for triggering and viewing dispute replays in this invention.

[0027] Figure 4 This is a schematic diagram of the time synchronization scheme of the present invention.

[0028] Figure 5 This is a schematic diagram illustrating the cropping range for different editing rules of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0030] This invention aims to provide a method and system for automatically generating video clips based on match score data, in order to solve the following technical problems: how to automatically identify exciting moments in a match without relying on video content analysis; how to achieve high-precision and high-efficiency automatic video editing using existing score data; and how to provide low-cost and easy-to-deploy video service capabilities for amateur table tennis events.

[0031] The purpose of this invention is to provide a method and system for automatically generating video clips based on match score data. By making full use of the score data (including the time and status of each point) collected in real time by the intelligent scoring terminal, the system uses it as a "time index" for video editing. It automatically identifies various exciting moments such as the whole match, each point, winning points, key points, comebacks, and controversial replays, and extracts the corresponding video clips. This achieves efficient and accurate automatic video editing without human intervention or AI visual analysis.

[0032] To achieve the above objectives, the present invention provides the following technical solution: a method for automatically generating video clips based on match score data, comprising the following steps: S1: Data Acquisition. Real-time acquisition of score data generated by the intelligent scoring terminal during sports events. The score data includes the time of each point and the score status at that moment. Simultaneously, acquisition of video stream data synchronized with the score data.

[0033] S2: Time synchronization, which synchronizes the time of the intelligent scoring terminal with the time of the video acquisition device to the same time base, ensuring that the timestamps of the score data and the video frames can correspond accurately.

[0034] S3: Rule matching. Based on preset editing rules, identify the target moment from the score data. Different editing rules correspond to different identification logic.

[0035] S4: Video extraction. Based on the identified target time, extract video segments within the corresponding time range from the video stream data to generate the target video segment.

[0036] S5: Tagging and storage. Tag the generated target video clips, identify their corresponding event types (such as "whole game", "key point", "winning point", etc.), and push them to the cloud or local storage for users to access via mobile devices.

[0037] Editing rules: Editing is driven by score data. The preset editing rules include the following types: Throughout the match, the starting point is identified when both players take their positions and trigger their actions, and the ending point is identified when the match ends. A preset duration is extracted from the moment the players take their positions to the moment the match ends. The application scenario is a complete match replay. Each playback segment identifies the moment each point is captured; a preset duration is extracted before and after the point's occurrence; the application scenario is independent playback of each point. Winning Score Highlights: Identifies the winning score that ends a game; extracts a preset time interval before and after the winning score occurs; applicable scenario: a collection of player wins; Key scores identify each point after the score reaches a preset key score value (such as 10:10, 12:12, etc.); extract a preset time before and after the point; the application scenario is key score highlights; The ability to identify a comeback from behind to take the lead; to extract a preset time interval from the start of the comeback to the end of the comeback; and to be used for highlight moments. Dispute replay responds to a dispute replay request triggered by a user via a mobile device. Based on the selected replay range, it extracts video clips corresponding to the current minute, the previous minute, or the two minutes prior. The application scenario is on-site dispute verification.

[0038] The preset duration can be flexibly configured according to actual needs, such as 1 second, 2 seconds, 3 seconds, 5 seconds or 10 seconds, to adapt to the playback needs in different scenarios.

[0039] Time synchronization technology: The time synchronization between score data and video stream is achieved through the following methods: Option A: NTP time synchronization. Connect both the intelligent scoring terminal and video acquisition equipment to an NTP time server to ensure all devices maintain consistent time. This method is simple, reliable, and suitable for locations with network coverage.

[0040] Option B: Timestamp Anchor Synchronization. Each time a minute is scored, the intelligent scoring terminal sends a synchronization signal containing a precise timestamp to the video acquisition device via wireless communication. The video acquisition device then establishes an index relationship between this timestamp and the current video frame, achieving frame-level synchronization.

[0041] Dispute replay process: For controversial balls that may occur in amateur competitions, a live replay function is provided through mobile devices, allowing players to view controversial moments via mobile phones and other mobile devices.

[0042] This invention also provides a video clip automatic generation system based on match score data, comprising: Intelligent scoring terminals: Located on both sides of the table tennis table, these terminals update the score in response to player strokes, recording the moment and status of each point. Two terminals are configured for each table (one for the home team and one for the away team).

[0043] Video capture equipment: Each table tennis table should be equipped with at least one network camera to capture video stream data of the table tennis area. The camera resolution should be sufficient to clearly identify the area of ​​the table tennis table; there is no specific limit to the value, and it can be flexibly selected based on venue conditions and cost.

[0044] Backend server: Communicates with the intelligent scoring terminal and video acquisition equipment, including: a data receiving module: receives score data and video stream data; an editing rule module: stores preset editing rules; a target time identification module: identifies the target time from the score data according to the rules; a video extraction module: extracts video segments according to the target time; a video storage module: stores the generated target video segments; and an access interface module: provides video segment access services to mobile devices. Mobile devices: Smartphones, tablets, and other devices used by players or spectators, running applications or mini-programs, used to: receive user-triggered replay requests for dispute replays; display replay options (current score, previous score, previous two scores); play dispute replay clips obtained from the backend server; and provide online viewing, downloading, and sharing functions for other video clips.

[0045] Example 1: A gym's monthly points-based tournament for members: In February 2026, the video automatic editing system based on score data, as described in this invention, was deployed for the first time in a monthly points tournament for members held by a table tennis club. The tournament consisted of 32 men's singles players, divided into 8 groups for a round-robin followed by a knockout stage. A total of 8 tables were used, each equipped with: 1 intelligent scoring terminal (including two button units on the home and away sides); 1 network camera (720P resolution, capable of clearly identifying the table area); players accessed the video service via a tournament mini-program on their mobile phones. Data acquisition and time synchronization: All cameras are connected to the local area network via a PoE switch to maintain NTP time synchronization with the backend server; the intelligent scoring terminal records a precise timestamp each time it takes a score and uploads it to the backend server via Wi-Fi. Editing rule configuration: The backend server presets the following editing rules, with the preset duration uniformly configured as 3 seconds: Throughout the match, the recognition logic is as follows: both sides take their positions and trigger the action → the match ends, capturing the period from 3 seconds before taking their positions to 3 seconds after the match ends. For each replay, the recognition logic is as follows: at each time a score is recorded, extract the 3 seconds before and 3 seconds after that score. The winning score highlights are identified by the following logic: for the last point of each round, extract the last 3 seconds before and after that point. The key score identification logic is as follows: for each score after 10:10, extract the data from 3 seconds before and 3 seconds after that score. The logic for identifying a comeback is as follows: from being down by 2 points to taking the lead, capture the period from the first 3 seconds of the comeback to the last 3 seconds of taking the lead. The logic for replaying disputes is as follows: users select the replay range via their mobile phones and extract the complete segment of the current score / previous score / previous two scores.

[0046] Actual Operation Results: Automatic Generation of the Entire Match: In the first match of Group A, Zhang San vs. Li Si. After both players took their positions, they confirmed their positions using their respective methods (clicking a button in this match), and the system recorded the moment of arrival. After the match ended (Li Si won 3-2), and the final point was recorded, the system automatically extracted the video from "3 seconds before taking position" to "3 seconds after the final point," generating a complete match video. The entire match video was generated within 3 seconds of the match ending. Automatic Generation of Key Point Highlights: In the final, the score was 10-10, and every point thereafter was automatically marked as a "key point." Zhang San ultimately won 13-11, and the system automatically generated a highlight video containing 3 key points, with a total duration of approximately 45 seconds. Zhang San immediately shared the highlight video on his WeChat Moments after the match, receiving numerous likes. On-site Application of Controversial Replays: In the semi-final, player Wang Wu's return shot was ruled a point for grazing the edge of the court, while his opponent Zhao Liu argued that the ball did not. Zhao Liu took out his phone, opened the tournament's mini-program, entered the current match page, and clicked the "Dispute Replay" button. The system displayed three options: "View Current Score," "View Previous Score," and "View Previous Two Scores." Zhao Liu selected "View Current Score," and the system immediately extracted the video clip corresponding to the current score and pushed it to Zhao Liu's phone for playback. After both sides watched the replay and confirmed that the ball had indeed grazed the side of the court, the match continued. The entire process took approximately 30 seconds, requiring no third-party intervention and no reliance on on-site display equipment.

[0047] A total of 32 video clips were generated during the event: 32 clips of the entire match, approximately 1,500 replays per point, 32 highlights of winning points, 86 clips of crucial points, 12 clips of comebacks, and 3 controversial replays. The total video storage was manageable (using H.265 encoding, with crucial point clips stored independently). Players could view and download their own match videos at any time via a mobile app, with a paid download conversion rate of approximately 15%. Sponsors purchased introductory advertisements for crucial point highlights and paid sponsorship fees.

[0048] Example 2: Multiple ways to trigger and view dispute replays: Editing is driven by score data, and dispute replays are primarily implemented via mobile devices, aligning with the practical usage scenarios of amateur tournaments. In other embodiments of this invention, the triggering and viewing methods for dispute replays may include: Mobile app / app interface operation: Users can click the replay option on the interface to select and view video clips of the current point, the previous point, or the previous two points. QR code viewing: A QR code is posted next to the table; players scan the code to enter the table's live page and click the replay button to view the replay. Push notification: After each point, the system automatically pushes a replay link to the mobile devices of the players in that match, which they can directly click to view. The reason for not recommending using physical buttons to trigger dispute replays on-site is that the button functions of the intelligent scoring terminal are already occupied (score recording, long-press to skip missed matches, etc.); adding physical buttons would increase hardware costs and complexity; mobile devices are readily available to everyone, offering convenient operation without requiring additional hardware investment.

[0049] Hardware selection guidelines: The resolution of the video capture equipment should meet the principle of "being able to clearly identify the range of the table tennis table," and can be flexibly selected based on venue conditions and budget. For example, the basic version uses a resolution of 480P / 720P, suitable for community stadiums and daily training. The standard version uses a resolution of 1080P, suitable for mid-to-high-end stadiums and member-only events. The flagship version uses a resolution of 4K, suitable for large-scale events and live streaming needs.

[0050] This invention is not only applicable to table tennis, but can also be adapted for use in badminton, tennis, volleyball, and other sports with clearly defined scoring rules and the use of scoring equipment to record each point. Any modification to the application scenario based on the technical solution defined in the claims of this invention, without departing from the core concept of this invention, should fall within the protection scope of this invention.

Claims

1. A method for automatically generating video clips based on match score data, characterized in that, include: S1. Real-time acquisition of score data generated by intelligent scoring terminals during sports events, the score data including the time of occurrence of each point and the score status at that time; S2. Obtain video stream data that is time-synchronized with the score data; S3. Identify the target moment from the score data according to the preset editing rules; S4. Based on the target time, extract video segments within the corresponding time range from the video stream data to generate the target video segment; S5, Tagging and Storage: Tag the generated target video clips to identify their corresponding event types and push them to the cloud or local storage for users to access via mobile devices.

2. The method for automatically generating video clips based on match score data according to claim 1, characterized in that: The preset editing rules include at least one of the following: The entire match rules are as follows: the first preset time before both players take their positions and trigger their actions is taken is taken as the start time, and the second preset time after the end time of the match is taken as the end time. The entire match video segment is extracted. Playback rules for each minute: Centered on the moment when each minute occurs, extract video segments of preset duration before and after that moment to generate an independent playback segment for each minute; the preset duration includes 1 second / 2 seconds / 3 seconds / 5 seconds / 10 seconds; Winning Point Highlights Rules: Identify all winning points that end a game, extract video clips of preset duration before and after each winning point, and generate winning point highlights; Key score rule: When the score reaches the preset key score value, extract video segments of preset duration before and after each subsequent point to generate key score segments; Comeback Rules: Identify the score change range from falling behind to taking the lead, extract a video segment from the start of the range to the end of the range with a preset duration, and generate a comeback segment; Dispute replay rules: In response to a dispute replay request triggered by a user via a mobile device, a video segment with the corresponding score is extracted from the video stream data according to the replay range selected by the user, and a dispute replay segment is generated; the replay range includes the current score, the previous score, or the previous two scores.

3. The method for automatically generating video clips based on match score data according to claim 2, characterized in that: The time synchronization between the score data and the video stream data is achieved in the following way: The time of the intelligent scoring terminal is synchronized with the time of the video acquisition device to the same time reference server; or the intelligent scoring terminal sends a synchronization signal containing a timestamp to the video acquisition device at each minute, and the video acquisition device establishes an index relationship between the timestamp and the video frame.

4. The method for automatically generating video clips based on match score data according to claim 1, characterized in that, Also includes: The generated target video clips are tagged with labels that identify the event type corresponding to the video clips. The event types include: the whole game, every point, winning point, key point, comeback, and controversial replay. It also includes: pushing the generated target video clips to cloud storage or local storage; providing an interface for users to access via mobile devices, supporting users to watch, download or share the video clips online; The intelligent scoring terminal includes trigger buttons located on both sides of the ping-pong table, which are used to respond to the player's hitting action to update the score and record the moment each point occurs.

5. The method for automatically generating video clips based on match score data according to claim 2, characterized in that: The actions to trigger the positioning of both players include clicking the button, double-clicking the button, pressing both buttons simultaneously, or confirming positioning via a mobile device. The end time of the match is determined in the following way: Once the last point of a game is detected and no new score data is available within a preset time period, the moment when the last point occurs will be taken as the end time of the game. The dispute replay request is triggered by an application or mini-program on a mobile device. Users can select to view the video clip of the current minute, the previous minute, or the previous two minutes by clicking the replay option on the interface.

6. A video clip automatic generation system based on match score data, characterized in that, include: At least one intelligent scoring terminal is used to collect score data in real time during a sports event, the score data including the time when each point occurs and the score status at that time; At least one video capture device is used to capture video stream data that is time-synchronized with the score data; At least one back-end server is communicatively connected to the intelligent scoring terminal and video acquisition equipment to receive score data and video stream data; The backend server includes: The data receiving module is used to receive score data and video stream data; The editing rules module is used to store preset editing rules; The target moment identification module is used to identify the target moment from the score data according to preset editing rules; The video extraction module is used to extract video segments within a corresponding time range from the video stream data according to the target time. The video storage module is used to store the generated target video segments; The access interface module is used to provide video clip access services to mobile devices.

7. The automatic video clip generation system based on match score data as described in claim 6, characterized in that: It also includes mobile devices, which run applications or mini-programs to receive dispute replay requests triggered by users, display replay options interfaces, and play dispute replay clips obtained from the backend server.

8. The automatic video clip generation system based on match score data as described in claim 6, characterized in that: The resolution of the video capture device is sufficient to clearly identify the range of the table tennis table.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the method as described in any one of claims 1-5.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-5.

Images