II – Project presentation

          2-1 The Projects’ Interest

          Despite all technological advances which have been made information retrieval faster and easier than ever before, this invaluable resource remains scattered, difficult to integrate, hard to locate. This is particularly the case for graphic and picture information. There is almost no technological tool for image information retrieval for instance for the research of an object represented on a picture that is retrieval by sample or content.

          Finding a lost image information image or represented objects in a picture on the Web, is a process that relies upon indexing structures which span a vast number of resources in many different sites all over the world.
However, the effectiveness of these indexes is highly dependent on the way these resources have been described by their providers.
Meta-data are “information” that make data useful and represents data management meant system catalogues or repository descriptors content dependent.
The concept of meta- data allows description of available resources . These are well developed on the Internet for textual information but much less for information about images. This project consists of filling out this gap.

          The meta-data concept supports a variety of functions that users have come to expect from search mechanisms such as:

For textual information meta-data standards have been proposed on the Internet:

     2-2 Difficulties we have encountered

          2-2.1- Due to heterogeneity

          2-2.2 In research mechanism

          Electronic documents in the HTML format represent most of the objects
accessed via www, generally without any associated extra information to improve their retrieval. The documents depend exclusively on the key- words and full text indexing used by the search tools without factor .
From where it appears:

     2-3 Problem analysis

     The adoption of information technologies in the whole range of human activities, as part of it’s effect, has developed the creation of more and more important volume data.

     Signals and digital images invade our current society, yet without means to adequately process them and then the facilities to easily manipulate them once they have been achieved.

     As the power and the richness of networks becomes ubiquitous, standard methods for accessing information and, more important, the implementation of object Oriented Methods in Information Classification, have prompted the selection of the Common Project Request Broker Architecture(COBRA) as the nucleus of an information object processing, classification, storage and retrieval system.

     The object of this document is to propose a new point of view for the PACHA with a view to address application level issues in terms of novel technique in data storage , management and the retrieval is based on a COBRA federated system. The primary objective is to concretise the work already achieved in the an industry ready product and to promote the results of the current PACHA architecture and systems.

     2-4 Rational development

          2-4.1-Proposed methodological approach

     Overall objective: The implementation for a system for massive digital storage available for consulting using content criteria.

     The complement system is planned for having two principal founding concepts as basis for a division of operational tasks.

     Firstly, this kind of data processing where incoming raw signals or image data is prepared for storage :consultation. The pre-processing results should provide sufficient information to enable strong characteristics and features to be extracted and stored. Thus satisfying the needs of interrogation.

     For the second phase of interrogation and the retrieval the content and data similarities will drive the selection of information selected and returned to the user for presentation.