Based On The Mpeg-7 Image Description And Retrieval

With the development of computer and communication technology, Multimedia data (such as video, audio and image) is increasing in a greatly increasing speed. If we cannot timely deal well with these data, we will have to facing more and more "Information Garbage". So it is obviously important for us that how to store and search this huge data more efficiently.The aim of the research of MPEG-7 is creating a kind of commonly description interface standard, which can satisfy the real-time (or not) requirement coming from various multimedia systems. The standard of MPEG-7 defined a great deal of data structure using to describe multimedia data. It can not only describe the low class characters (such as video and audio characters), but also it can describe the high-class characters (such as semantic, object and structure characters).But there are many MPEG-7 self-defined description schemes in the standard of MPEG-7, these description schemes are defined with special language, which usually named DDL. The DDL is extended from XML. So as a developer who wants to use MPEG-7 to describe multimedia content, you must have a deeply understanding about MPEG-7 DDL and MPEG-7 MDS.In this paper, we propose a description scheme of multimedia content, which is based on the standard of MPEG-7, especially designed for image content. And then based on this, we propose a searching engine for image content. In .details the research of this paper have the following four points.1. Image content description based on the standard of MPEG-7In this paper, we proposed a new image description scheme based on the standard of MPEG-7. This scheme directly uses XML to describe image content (the standard of MPEG-7 uses DDL, which is a description language based on XML). What is defined in MPEG-7 MDS is mainly included in our description scheme. Further more, it is easy to convert between the description and MPEG-7 description.2. Object relation graph and object hierarchy graphIn the scheme we proposed, we add Object relation graph and object hierarchy graph, so it can represent image semantic contents more sufficiently. You can define any relations and hierarchies you want, and then search them.3. Directly searching XML filesIn the searching engine we proposed, it can directly search XML files which include image contents. We parse the XML files beforehand to create indexes of content, and then search the indexes.4. Image searching engine that integrates several kinds of image charactersThe searching engine can not only realize semantic-based image searching with Chinese words, but also can realize more clever searching through integrating several different kinds of image characters, such as image semantic character, image color character and image using history. We can also extend this engine to realize object-oriented multi-level abstraction searching.