【作者】 谭罗生；

【导师】 黄明和； 聂承启；

【作者基本信息】 江西师范大学 ， 计算机应用， 2004， 硕士

【摘要】 进年来多媒体数据与日剧增，多媒体访问技术也日渐丰富。然而实际应用中要想从这些海量数据中搜寻感兴趣的信息并利用这些有用信息却依然是公认的难题。对于文本，现在已经有了很多搜索引擎来为用户搜索文本信息，但对于视频、音频等内容，却无法通过现有的搜索引擎查询到。因此如何快速有效地搜索感兴趣的各种多媒体素材已成为一大挑战。 MPEG-7将确立各种类型的多媒体信息的描述方法，这种描述与内容密切相关，并支持对用户感兴趣材料的快速和高效地检索与定位。未来的多媒体服务需要提供基于内容的快速和有效的检索服务，这其中MPEG-7技术将发挥多方面的作用，如提供对多媒体数据对象的内容和语义进行精确描述的元数据，采用Xlink链结机制实现描述数据与媒体数据的链接，使得从内容的描述能够快速地存取相关的媒体数据，扩展现在对多媒体的有限查询能力。 本文对实现基于MPEG-7标准的内容描述和描述生成及其在视频检索中的应用进行了比较全面的探索。主要工作概括如下：(1)对多媒体数据内容的元数据表示、组织和索引，多媒体数据的特征提取，多媒体内容的结构化及多媒体数据描述模型进行了详细分析，着重研究解决多媒体信息的组织、如何表示、内容的元数据描述、描述生成、使用等方面的关键性技术问题，其中，元数据的组织包含元数据模式、模式的名称空间和元数据与被描述数据内容间的链接。(2)对视频内容的MPEG-7描述及描述的生成进行了探讨。采用树状层次结构模型来表示描述文档，通过构造一棵K叉树(K是树中的最大分支数)的来生成MPEG-7描述。(3)设计了一个基于MPEG-7标准的视频描述生成和检索原型系统，并实现了其中的主要功能。更多还原

【Abstract】 In recent years, there has been a major increase in available multimedia and in technologies to access the multimedia.However, the extraction of useful information from multimedia and the application of this information in practical systems are still open problems. Now,many text information search enginges are available,but all of them can’t sustain multimedia such as A/V data yet.Therefore, how to search quickly and effectively the interested multimedia matelerials has been a challenging problem.MPEG-7 (Multimedia Content Description Interface) will establish standard descriptive method of kinds of multimedia information,the description has tight relation with information content,and will support rapid and effective access to material which will attract the user. It requires rapid and effective retrieval services based on content, efficient data organization and correlative materials quickly location.For this reason,MPEG-7 technology may play an important role in many aspects, such as supplying metadata to describe contents and semantic meaning of multimedia, accomplishing the accession to materials through Xlink and expanding the capacity of search of information from multimedia.In this paper,the application of MPEG-7 in content description generation and information retrieval of video has been systematically studied.The main work can be summarized as the following: First, the paper makes a detailed research on the metadata’s representation, organization and index of multimedia contents,features retrieval of multimedia data and description model of multimedia contents.We aims at to research the key technological problem of the organization and the representation of multimedia contents,the multimedia contents description ,the produce and using of description etc.The organization of metadata includes many technologies such as metadata schema,schema’s namespaces and linking mechanism that associates description and collerative multimedia materials. Second,the paper workes over the generation of MPEG-7 description,it’s core idea is through adopting hierarchy model to organize description materials,to generates MPEG-7 description with constructing a k-tree.Third, the paper has designed a prototype system of video content description generation and information retrieval based on MPEG-7 and realizing the major functions of the system.更多还原