KR20100070952A - Multimedia content management system - Google Patents

Abstract

  The present invention relates to securing the technology necessary for multimedia content processing and analyzing the module.

  The purpose of the present invention is to secure a technology necessary for multimedia content processing, and to analyze and design modules and to develop a multimedia content management system incorporating the technologies of designed modules. .

  The present invention for achieving the above object is largely divided into three modules of the developed multimedia management system video analysis module, clip information analysis module, clip information registration module. In addition, first, the video analysis module is divided into basic information extraction function, key frame extraction function, and clip generation function. Second, clip information analysis module is divided into emotional information analysis function and content information analysis function. Lose. Finally, the clip information registration module is divided into metadata registration function, MPEG-7 analysis function, and database registration function. The integrated module features content processing management techniques.

Description

Multimedia Content Management System {Multimedia Content Management System}

Broadcast technology field

   1. Video Analysis Module

  The video analysis module in the multimedia content management system is largely divided into basic video extraction function, first frame extraction function, and clip generation function. This module checks basic information about video in video basic information extraction function of video analysis module and extracts first frame and creates clip.

end. Video basic information extraction function

  In the basic video extraction function, the basic video production and file information is displayed to the user. In general, all video files have some form of information, not only videos provided on the web, but also videos taken by individual camcorders and video cameras, and include file name, date of creation, size, video type, and total video time. consist of. This information was used as metadata information of the clip created when the clip was created.

I. First frame extraction function

In the first frame extraction function, the first frame extraction of the clip by converting the video is the key in order for the user to create a video clip. The frame, which is the minimum unit that constitutes a video, is a still image, which becomes a video by continuous motion as the frame progresses. In the key frame extraction function, the first shot is extracted as the first shot of the frame in which the amount of change of information is greatly generated by comparing the frames by using the characteristic information obtained for each frame of the image. The whole video is divided into several episodes, and the episode can be split into several shots. Frames do not have time differences and have similar feature information. In the developed module, the difference between the first frame and the second frame is numerically compared using the Peak Signal to Noise Ratio (PSNR) technique, which is widely used in image processing for the first shot extraction.

All. Clip creation function

In the clip creation function, the user can create a clip by using the first shot extracted. Check the clip information on the clip and create the clip. The clip's information contains the basic movie information and various information for creating a database. In the video analysis module, the user can check the clip created by the first shot extraction function in the clip view and clip information.

    2. Clip Information Analysis Module

The clip information analysis module in the multimedia content management system is divided into emotional information analysis function and content information analysis function. The emotion information analysis function extracts color emotion information of an image, and the content information extraction function automatically extracts feature information about color and motion.

      end. Emotional information analysis function

In order to search and manage the image applying emotion, the emotional elements (color, shape, texture, and pattern) of the image were identified, and the emotional elements and the emotional vocabulary were matched to one space by using factor analysis among statistical techniques. By concordance analysis, we extracted verification of factor analysis and representative emotional vocabulary.

      I. Content based analytics

In the content-based analysis function, feature information is automatically extracted and input into MPEG-7 by color information extraction, texture information extraction, shape information extraction, and motion information extraction from the first shot of the generated clip and the entire frame of the clip. Modules for extracting information based on content are designed and used.

    3. Clip information registration module

The clip information registration module in the multimedia content management system is largely divided into a metadata registration function, an MPEG-7 analysis function, and a database registration function. In the clip information registration module, the metadata registration function creates the final information through the database registration function after re-analyzing the user's basic information on the created clip and the information to be added and deleted in the MPEG-7 analysis function. A module that stores a database of captured clips.

      end. Metadata registration function

To describe metadata, grasp the structure of video information to be described, determine the structural unit of video that is the basis of metadata system, identify what information is composed, and identify the data elements of metadata by these physical structural units The information is described for each structural unit. In the present invention, the metadata information is described for each clip using the clip generated by the video analysis module as a structural unit.

In addition to such a physical structure, in order to select an appropriate data element of video information to be used for metadata, it is necessary to understand the composition of the video information. In other words, it is to select metadata elements that can be expressed by identifying what elements the information is specifically configured and what elements are necessary for video information technology. In the present invention, the video information consists of bibliographic information, visual information, and semantic information.

      I. MPEG-7 creation and editing

In the MPEG-7 creation and editing function, metadata can be described and edited by the user using MPEG-7 for efficient retrieval of image data from a multimedia database. MPEG-7, which is used to store multimedia data attribute values, is mainly used as a technical solution for audio and visual information (still image, video, etc.).

In the present invention, it is an MPEG-7 editing tool developed for efficient generation and management of MPEG-7 data used for representing the main attributes of multimedia. Tree-type structural browser is available for MPEG-7 file, and insert and delete function of each node is added to manage metadata efficiently. The created MPEG-7 file shows the entire structure in metadata in XML format in the metadata source view function. In addition, it includes simple functions such as grammar check for correcting wrong grammar when writing metadata.

      All. Database registration function

In the database registration function, users can register and upload clip information created by the clip information analysis module, metadata registration function, and MPEG-7 editing tool using My SQL. In addition, using the information stored in the database registration function, such as keywords, emotions, content (color, shape, motion) can be searched to find the clips desired by the emotion-based intelligent multimedia search system.

  Development of Multimedia Contents Data Structure Design and Analysis Management Tool; Data structure design based on feature analysis and feature analysis of video and audio; Building various multimedia contents database; Exchange through agreements with broadcasting stations and video production companies; Image segmentation of video data; Study on Extraction Method of Semantic Representative Frame of Video Image; Development of Semantic Image Feature Extraction Techniques for Multimedia Contents; Semantic unit image segmentation by developing video feature information extraction automation system;

  Focusing on developing the base system of multimedia contents management system technology, we performed the semantic analysis technology based on the characteristics of multimedia data and the invention of intelligent multimedia processing. In particular, it focused on the development of the structure design and analysis management tool and the semantic feature extraction technique of multimedia contents data.

  By developing a multimedia content management system, it is possible to manage contents systematically and efficiently. Providing intelligent contents production and editing function based on multimedia characteristics elements, it provides convenience not only to general users but also experts, and provides intelligent and meaningful services for emotional needs of users. Different from service. As a plug-in of multimedia contents (video, music, character, etc.) production systems, this system can be added and used as an analysis system to extract feature elements. Developing a search and production system based on a new concept of content and emotional information, strengthening national technological competitiveness in the global era.

Design of multimedia database structure and development of system to manage it

 ■ Development of multimedia content data structure design and analysis management tool

 ■ Build various multimedia contents database

 ■ Image segmentation method development of video data

 ■ Development of Semantic Image Feature Extraction Technology for Multimedia Contents

(1) Image content data structure design and analysis management tool development

 ■ Additional development and integrated management of complex query and extraction techniques of descriptors and features

 ■ Increase the accuracy of existing designed algorithms and improve search throughput.

 ■ Extraction of feature descriptors invariant to movement, rotation, contraction, and expansion in algorithms using color, image space, and form

 ■ Development of user-selective combination technology that can combine color, texture, and shape descriptors described in MPEG-7 to extract feature information of video

 ■ Modification of multimedia content search engine to retrieve XML-based data for database after algorithm design of feature elements

(2) Constructing various multimedia content databases

 ■ Multimedia feature based database improvement and improvement

(3) Development of video segmentation technique of video data

 ■ Implementation of Extraction Techniques for Representative Video Frames

 ■ Semantic Unit Image Segmentation and First Frame Extraction of Video Data

 ■ Shot using characteristic information of video information (meaning clip of video in broadcasting term)

 ■ Development of automated image extraction system

-Development of various extraction algorithms of representative frame extraction technology

-Development of automation system such as object classification, identification, and flow of shot image information

-Development of the method of organizing the semantic information of shots corresponding to the representative frame into metadata

(4) Development of semantic image feature extraction technology for multimedia contents

 ■ Development of weighting technology of MPEG-7 descriptor suitable for video extracted from video images (various contents, etc.)

 ■ Development of search technology combining semantic keyword-based method which improved search method using simple keyword-based method

1 is a multimedia content management system

2 is a video basic information function of the video analysis module

3 is a first shot extraction function of the video analysis module

4 is a clip generation function of the video analysis module

5 is an emotion-based analysis function of the clip information analysis module

6 is a content-based analysis function of the clip information analysis module

7 is a metadata registration function of a clip information registration module

8 is a function for generating and editing MPEG-7 of a clip information registration module.

9 is a database registration function of the clip information registration module.

10 is a flowchart of a multimedia content management system

Claims (3)

 In a method for analyzing a video in a multimedia content management system, (FIG. 1)  The first shot is a frame in which a large amount of change of information is generated through frame-by-frame comparison using the basic video extraction function (FIG. 2), which shows the file information to the user, and feature information obtained for each frame of the video. The first frame extraction function that can extract (Fig. 3), the user can create a clip using the first shot extracted, check the clip information for the clip, create a clip and the information of the clip Clip creation function with basic information and various information for creating a database (Fig. 4)  In the method for analyzing clip information in a multimedia content management system,  Emotion information analysis function (FIG. 5) to grasp the emotion elements (color, shape, texture, pattern) that the image has for the management of the applied emotion, extract the color information from the first shot of the generated clip and the entire frame of the clip, Content-based analysis technology in which feature information is automatically extracted by texture information extraction, shape information extraction, and motion information extraction (FIG. 6)   In the method for registering the clip information in the multimedia content management system,  In order to select an appropriate data element of the video information to be used for the metadata, a metadata registration function (FIG. 7) that identifies and registers the configuration of the video information, and the user uses MPEG-7 for efficient retrieval of the image data from the multimedia database. MPEG-7 creation and editing function (FIG. 8) that allows metadata description and editing, clip information analysis module, metadata registration function, and clip information generated by MPEG-7 editing tool using My SQL. Database registration function that can register and upload (Fig. 9)

Images