Design And Implementation Of Multimedia Information Retrieval Client System

Multimedia information concern technologies in many different fields, Such as database management system, digital signal processing, image processing and analysis, optical communication, mass storage, artificial intelligence, new scheme of programming and text processing etc. The combination of these issues is key to the development of multimedia information system.Multimedia information systems need the supports of traditional DBMS. After these years of developing, traditional DBMS have formed a complete specification and developing model. This provides a template for the case of multimedia We can study from it in the following area: data abstract, high level query language, system neutrality, and the architecture. However we demand more powerful performance for it than its ancestor, Because of the heterogeneous nature of multimedia data.The system should have great adjustment in data storing, data transporting, data presentation and management.Traditional DBMS is only suited for the processing of simple data like digital number, character and string. It provides nothing for the processing of multimedia information.Multimedia information system must provide a lot of media concerning operations, such as character refining, indexing, calculating.And new problems including media combination and media presentation should be dealt in the client side.We developed a multimedia information system called MIRES-Multimedia Information Retrieving System, which provide content based retrieving services. It is a distributed computing system based on Browser-Server architecture.We exploit the newly developed JAVA technology to provides service with zero installation costs in client side. A client can access the MIRES using a general browser like NETSCAPE or INTERNET EXPLORER.The MIRES currently provides full text retrieval and image retrieval. To use full text retrieval, users just submit the query statement to the server, and can get the result with marked fitness. In the case of image retrieval, we provide a toolbox in client side,which help users expressing the characteristics of the image they want to get.Through this way users can do the job in a visual way.