Abstract: With the amount of superhighways, a vast amount of data is currently available on the internet. The advances in the methods for processing and coding images, video and audio have led to a wide spectrum of non-traditional forms of data being available across the network. Text, images video and audio data, all perhaps stored in multiple heterogeneous formats, form the core of what is known today as multimedia data. 3D objects are an important multimedia data type with many application possibilities. 3D models can represent complex information, and the problem of content-based searching in large 3D object repositories arises in a number of practical fields. The large amounts of 3 D objects as multimedia data that are being collected in multimedia archives have originated the need for means to search the archives. It is recognized that, unlike text, for which effective content-based retrieval ways are presently used, it is challenging to automatically indexed or search 3D image supported their content. For this, it is necessary to make use of any available description associated with the objects, concerning either the content itself or the context of the objects’ production and use. Metadata is getting a fair share of attention in current research on solutions to the querying of multimedia objects. But retrieval of multimedia information on the basis of only keyword or metadata will not solve the problem, in the present manuscript we have proposed a web based multimedia information storage and retrieval model is proposed for the integration of metadata and content. This model is tested on the implementation of a prototype of multimedia information storage and retrieval system based on a relational database. Statistical functions are used on medical images to calculate similarity for storage and retrieval.

Keywords: 3D images, Multimedia, Visual Similarity, Storage and Retrieval.