Multimedia data analysis method and client
By screenshotting and subject analysis of multimedia data on the client and identifying and identifying the main part, the problem of excessive resource utilization on the server is solved, and more efficient multimedia data analysis is achieved, and the overall efficiency and analysis accuracy of the system are improved.
Patent Information
- Application Number
- CN202411018650.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-29
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2044-07-29
AI Technical Summary
In the prior art, the analysis of video streams or pictures requires a large amount of calculations on the server, resulting in an increase in resource usage on the server, a double pressure, and affecting system efficiency.
Screenshots and subject analysis of multimedia data are carried out on the client, and the main part is identified and identified, the number of times is summarized, and the results are sent to the server. The server model is used for further analysis to reduce the amount of data transmitted and avoid occupying server resources.
By conducting preliminary analysis on the client, we can reduce the amount of data processing on the server, reduce resource usage, improve analysis accuracy and overall system efficiency, and reduce the pressure on the server.
Smart Images

Figure CN119011904A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data analysis, and in particular to a multimedia data analysis method and a client. Background Art
[0002] At present, there are many image recognition functions or video stream key information recognition functions on the market, such as video scanning content, video stream related recognition and key content marking functions. The main principle is that the client generates a video stream and transmits it to the server. The server uses a large model to perform data analysis on the video stream. The server needs to analyze all video streams or pictures, and the pressure on the server is doubled. Therefore, there is an urgent need for a multimedia data analysis method and client that can solve the above problems. Summary of the invention
[0003] The technical problem to be solved by the present invention is to provide a multimedia data analysis method and client, which can avoid occupying server resources, relieve server pressure, thereby speeding up server calculation and analysis, and further improving the overall efficiency of the system.
[0004] In order to solve the above technical problems, the technical solution adopted by the present invention is:
[0005] A method for analyzing multimedia data, comprising the steps of:
[0006] S1. Acquire a multimedia file to be analyzed, take screenshots of the multimedia file at a preset frequency, and perform a main body analysis on the screenshots;
[0007] S2. Send each frame screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to analyze the received data and output the analysis results.
[0008] In order to solve the above technical problems, another technical solution adopted by the present invention is:
[0009] A multimedia data analysis client comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, each step of the multimedia data analysis method described above is implemented.
[0010] The beneficial effects of the present invention are as follows: the present invention provides a multimedia data analysis method and client, obtains a multimedia file to be analyzed, takes a screenshot of the multimedia file at a preset frequency, performs a main body analysis on the screenshot, identifies the main body area of the screenshot and sets an identifier for the identified main body part, summarizes the total number of times each main body part has been identified, which helps to improve the accuracy of subsequent server-side model recognition and analysis, and sends the analysis results and the multimedia file to the server as input parameters of the server-side model, uses the server-side model for further analysis, makes full use of the server-side resources, improves the depth and breadth of the analysis, and outputs more accurate analysis results, and effectively reduces the amount of data transmitted to the server by performing recognition and preliminary analysis on the client side, avoids occupying server-side resources, reduces server-side pressure, speeds up server-side calculation and analysis, and further improves the overall efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] Figure 1 A flowchart of a method for analyzing multimedia data according to an embodiment of the present invention;
[0012] Figure 2 A schematic diagram of a multimedia data analysis client according to an embodiment of the present invention;
[0013] Description of labels:
[0014] 1. A multimedia data analysis client; 2. A memory; 3. A processor. DETAILED DESCRIPTION
[0015] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.
[0016] Please refer to Figure 1 , an embodiment of the present invention provides a method for analyzing multimedia data, comprising the steps of:
[0017] S1. Acquire a multimedia file to be analyzed, take screenshots of the multimedia file at a preset frequency, and perform a main body analysis on the screenshots;
[0018] S2. Send each frame screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to perform analysis and output the analysis results.
[0019] From the above description, it can be seen that the beneficial effects of the present invention are: obtaining a multimedia file to be analyzed, taking a screenshot of the multimedia file at a preset frequency, performing a main body analysis on the screenshot, identifying the main body area of the screenshot and setting an identifier for the identified main body, which helps to improve the accuracy of subsequent server-side model recognition and analysis, and sending the analysis results and the multimedia file to the server as input parameters of the server-side model, using the server-side model for further analysis, making full use of the server-side resources, improving the depth and breadth of the analysis, and outputting more accurate analysis results, and effectively reducing the amount of data transmitted to the server by performing identification and preliminary analysis on the client side, avoiding occupying server-side resources, reducing server-side pressure, and speeding up server-side calculation and analysis, thereby further improving the overall efficiency of the system.
[0020] Furthermore, taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots includes:
[0021] The type of the multimedia file is determined. If the type of the multimedia file is a video stream, frame extraction and screenshots are performed on the video stream at a preset frequency, and a main body analysis is performed on the screenshots.
[0022] From the above description, it can be seen that different processing methods are adopted according to different types of multimedia files. When facing video-type multimedia files, frame screenshots are taken at a preset frequency, and the video is converted into pictures for identification and analysis. This not only ensures the integrity of the multimedia data, but also reduces the amount of data processed and improves processing efficiency. In addition, this conversion can more clearly display the key information in the video, reduce analysis errors caused by problems such as video playback speed or frame rate, and help improve the accuracy of data analysis.
[0023] Furthermore, the preset frequency is dynamically set according to the video length. If the video length does not exceed the preset length, the video stream is framed and screenshotted at the preset frequency. If the video length exceeds the preset length, the video length and the preset length are used to calculate the excess ratio of the length, the preset frequency is reduced according to the excess ratio, and the video stream is framed and screenshotted at the reduced frequency.
[0024] From the above description, it can be seen that dynamic settings are made according to the length of the video. If the video length does not exceed the preset length, the preset frequency is used for frame extraction and screenshot. Otherwise, the preset frequency is reduced and frame extraction and screenshot are performed. This dynamic setting can adapt to the uncertainty of the video length. For long videos, the preset frequency is reduced, the screenshot interval is increased, the number of screenshots is reduced, the amount of data processed is reduced, and the efficiency of data analysis is improved.
[0025] Furthermore, the operation of performing the main body analysis on the screenshot is specifically: identifying the main body area of the screenshot and setting a mark for the identified main body part, and summarizing the total number of times each main body part has been identified.
[0026] From the above description, it can be seen that by identifying the main area of the screenshot, it is ensured that the main part can be accurately and quickly identified, and an identification is set for the identified main part to facilitate subsequent analysis and statistics, summarize the total number of times each main part has been identified, and finally determine the main body of the picture, thereby realizing the recognition and main analysis of multimedia files, and laying the foundation for the subsequent server to use a large model for analysis and output more accurate results.
[0027] Furthermore, the taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots further includes:
[0028] The type of the multimedia file is determined. If the type of the multimedia file is a picture, the picture is directly subjected to a main body analysis.
[0029] From the above description, it can be seen that different processing methods are adopted according to different types of multimedia files. When facing multimedia files of picture type, data analysis is performed directly to avoid unnecessary frame extraction and screenshot processing, reduce resource waste, and realize rapid identification and subject analysis of pictures to obtain analysis results.
[0030] Please refer to Figure 2 Another embodiment of the present invention provides an analysis client for multimedia data, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements each step of the above-mentioned multimedia data analysis method when executing the computer program.
[0031] The above-mentioned multimedia data analysis method and terminal of the present invention are suitable for multimedia data analysis, and are described below through specific implementation methods:
[0032] Embodiment 1
[0033] Please refer to Figure 1 , a method for analyzing multimedia data, comprising the steps of:
[0034] S1. Acquire a multimedia file to be analyzed, take screenshots of the multimedia file at a preset frequency, and perform a main body analysis on the screenshots.
[0035] In this embodiment, screenshots of multimedia files are taken at a preset frequency, which effectively reduces the amount of data to be processed, significantly improves the processing speed, reduces computing resource consumption, and extracts screenshots as key frames of multimedia content to facilitate subsequent subject analysis and content recognition, making the entire analysis process more efficient.
[0036] Furthermore, in the present embodiment, taking screenshots of the multimedia file at a preset frequency and performing a main body analysis on the screenshots include: determining the type of the multimedia file; if the multimedia file is a video stream, taking frame screenshots of the video stream at a preset frequency and performing a main body analysis on the screenshots; if the multimedia file is a picture, directly performing a main body analysis on the picture; and taking different processing methods according to different types of multimedia files. For multimedia files of video type, the frequency is dynamically adjusted according to the length of the video, frame screenshots are taken, and the video is converted into a picture for identification and analysis, which not only ensures the integrity of the multimedia data, but also reduces the amount of data processed, improves processing efficiency, and helps to improve the accuracy of data analysis. For multimedia files of picture type, data analysis is directly performed to avoid unnecessary frame screenshot processing, reduce resource waste, and enable rapid identification and preliminary analysis of the picture to obtain analysis results.
[0037] Furthermore, in the present embodiment, the preset frequency is dynamically set according to the video length. If the video length does not exceed the preset length, the video stream is framed and captured at the preset frequency. If the video length exceeds the preset length, the video length and the preset length are used to calculate the excess ratio of the length. The frequency is reduced according to the excess ratio, and the video stream is framed and captured at the reduced frequency. By dynamically setting the video capture frequency to adapt to the uncertainty of the video length, for long videos, the preset frequency is reduced, the screenshot interval is increased, the number of screenshots is reduced, the amount of data processed is reduced, and the efficiency of data analysis is improved.
[0038] Furthermore, in the present embodiment, the operation of performing the main body analysis on the screenshot is specifically as follows: identifying the main body area of the screenshot and setting an identification for the identified main body part, summarizing the total number of times each main body part has been identified, ensuring that the main body part can be accurately and quickly identified, so as to facilitate subsequent analysis and statistics, and finally determine the main body of the picture, thereby realizing the recognition and preliminary analysis of multimedia files, and laying the foundation for the subsequent server to adopt a large model for analysis and output more accurate results.
[0039] S2. Send each frame screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to analyze the received data and output the analysis results.
[0040] In this embodiment, the analysis results and multimedia files are sent to the server as input parameters of the server model, and the server model is used for further analysis. For example, a video of a cat playing with a duration of 100 seconds is framed and screenshotted every 5 seconds, with a total of 20 frame screenshots, and the main body analysis is performed on these 20 screenshots. The main body of each screenshot is different, and the main body includes a cat, grass, a table and a house. The identified main body is marked and summarized to obtain data on the total number of times each main body is identified: cat=20, table=3, house=2, grass=5. Finally, the 20 screenshots, the main body, the total number of times each main body and the video are sent to the server, so as to make full use of the resources of the server, improve the depth and breadth of the analysis, and output a more accurate analysis result. By performing identification and main body analysis on the client side, the amount of data transmitted to the server is effectively reduced, the server resources are avoided from being occupied, the pressure on the server is reduced, and the calculation and analysis on the server are accelerated, thereby further improving the overall efficiency of the system.
[0041] Embodiment 2
[0042] Please refer to Figure 2 A multimedia data analysis client 1 includes a memory 2, a processor 3, and a computer program stored in the memory 2 and executable on the processor 3. When the processor 3 executes the computer program, each step of a multimedia data analysis method of embodiment 1 is implemented.
[0043] In summary, the present invention provides a multimedia data analysis method and client, which obtains a multimedia file to be analyzed, takes screenshots of the multimedia file at a preset frequency, and determines the type of the multimedia file. If the type of the multimedia file is a video stream, the preset frequency is dynamically set according to the video duration. When the video duration does not exceed the preset duration, the video is framed and screenshoted at the preset frequency. When the video duration exceeds the preset duration, the video duration and the preset duration are used to calculate the excess ratio of the duration, the frequency is reduced according to the excess ratio, the video stream is framed and screenshoted at the reduced frequency, and the screenshots are subjected to main analysis. The video screenshot frequency is dynamically set to adapt to the uncertainty of the video duration, thereby improving the efficiency of data analysis. If the type is a picture, the main body analysis is performed directly on the picture. Different processing methods are adopted according to the types of different multimedia files, which can not only avoid resource loss, but also realize rapid identification and preliminary analysis of the main area of the screenshot, and set an identification for the identified main part, summarize the total number of identified main parts, and send each frame of screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to analyze the received data and output the analysis results. By identifying and analyzing the main body of the screenshot data on the client side, the amount of data transmitted to the server is effectively reduced, avoiding occupying server resources, alleviating server pressure, and speeding up server calculation and analysis, thereby further improving the overall efficiency of the system.
[0044] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.
Claims
1. A method for analyzing multimedia data, characterized in that: Includes steps: S1. Acquire a multimedia file to be analyzed, take screenshots of the multimedia file at a preset frequency, and perform a main body analysis on the screenshots; S2. Send each frame screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to analyze the received data and output the analysis results.
2. A multimedia data analysis method according to claim 1, characterized in that: The taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots includes: The type of the multimedia file is determined. If the type of the multimedia file is a video stream, frame extraction and screenshots are performed on the video stream at a preset frequency, and a main body analysis is performed on the screenshots.
3. A multimedia data analysis method according to claim 2, characterized in that: The preset frequency is dynamically set according to the video length. If the video length does not exceed the preset length, the video stream is framed and screenshotted at the preset frequency. If the video length exceeds the preset length, the video length and the preset length are used to calculate the excess ratio of the length. The preset frequency is reduced according to the excess ratio, and the video stream is framed and screenshotted at the reduced frequency.
4. The method for analyzing multimedia data according to claim 1, characterized in that: The operation of performing the main body analysis on the screenshot is specifically: identifying the main body area of the screenshot and setting a mark for the identified main body part, and summarizing the total number of times each main body part has been identified.
5. The method for analyzing multimedia data according to claim 1, characterized in that: The method of taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots further includes: The type of the multimedia file is determined. If the type of the multimedia file is a picture, the picture is directly subjected to a main body analysis.
6. A multimedia data analysis client, 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 computer program, the following steps are implemented: S1. Acquire a multimedia file to be analyzed, take screenshots of the multimedia file at a preset frequency, and perform a main body analysis on the screenshots; S2. Send each frame screenshot, the main part, the total number of each main part and the multimedia file to the server, so that the server uses the model to analyze the received data and output the analysis results.
7. The multimedia data analysis client according to claim 6, characterized in that: The taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots includes: The type of the multimedia file is determined. If the type of the multimedia file is a video stream, frame extraction and screenshots are performed on the video stream at a preset frequency, and a main body analysis is performed on the screenshots.
8. The multimedia data analysis client according to claim 7, characterized in that: The preset frequency is dynamically set according to the video length. If the video length does not exceed the preset length, the video stream is framed and screenshotted at the preset frequency. If the video length exceeds the preset length, the video length and the preset length are used to calculate the excess ratio of the length. The preset frequency is reduced according to the excess ratio, and the video stream is framed and screenshotted at the reduced frequency.
9. The multimedia data analysis client according to claim 6, characterized in that: The operation of performing the main body analysis on the screenshot is specifically: identifying the main body area of the screenshot and setting a mark for the identified main body part, and summarizing the total number of times each main body part has been identified.
10. The multimedia data analysis client according to claim 6, characterized in that: The method of taking screenshots of the multimedia file at a preset frequency and performing main body analysis on the screenshots further includes: The type of the multimedia file is determined. If the type of the multimedia file is a picture, the picture is directly subjected to a main body analysis.
Citation Information
Patent Citations
Video stream processing method and device, electronic equipment, storage medium and system
CN114255432A
Video edge preprocessing and playing method and system
CN115065806A
Multimedia information identification method and device and electronic equipment
CN117151967A
Image processing method and apparatus in video live streaming process, and storage medium
US20190364333A1
Video pushing method, video pushing device and storage medium
US20240251131A1
