FORMULA

Automatic Feature Extraction for Efficient Retrieval in Large Multimedia Database

European Commission

Information Technologies Programme
ESPRIT #22246
Human Comfort and Security for Software Intensive System Engineering

Partner / Customer :

• Giunti Multimedia (IT), and Kino (GR), provide the user requirements and validate the demonstrator
• Fraunhofer IGD provides the enabling technologies
• TXT (IT), Epsilon (GR), and Fraunhofer IGD develop the tool jointly

Duration :

7/96 to 3/97

Motivation

Image database systems with high performance retrieval facilities are a pivotal issue of any future attempt for the development of an efficient organisation, management, and utilisation the flow of information in the electronic publishing context. These systems will impact the productivity of its users, conditioning their way of working in terms of both quality and quantity.

Scope

The FORMULA project aims at globally improving the electronic publishing process with enhanced image management and remote image databases. The targeted user domain is electronic publishing (e.g. press, CD, video). Electronic publishers have to deal with heterogeneous multimedia data on a daily basis. The composition of their products requires the collection of data from huge and unstructured databases, either electronic or not, whose handling is everything but simple or easy. Even when the support of electronic database is given, the insertion and retrieval of data areindeed costly and complex tasks. As of now, this data management is definitely inefficient and unsatisfactory. These limitations cause an impressive wasting of resources in terms of time and money, and prevent a more optimised production process.

Rationale

The result of the project will be a demonstrator for image database management, proving the integrability of new technologies based on automatic extraction of image annotations (during insertion) and on a pictorial content-based approach (during retrieval), for an innovative approach to image handling in electronic publishing.

Technologies

It will adapt and integrate different technologies from various related domains: Human Computer Interfaces (HCI), Content-based Image Retrieval (CBIR), Multimedia Databases (MMDB), World Wide Web (WWW), Integrated System Digital Network (ISDN), and other technologies for handling the multimedia data according to the identified user needs.

Benefits

This demonstrator will grant a reduction of costs faced by electronic publishers caused by image insertion, image retrieval, and other administrative issues by:

• reducing the need of extensive textual annotation of images.
  
• reducing the image insertion procedure to a simple scanning operation.
  
• reducing the time spent for image retrieval, thanks to the higher precision and speed of the retrieval mechanism.

Consortium

The FORMULA consortium provides all the necessary expertise to achieve the above objectives and to successfully exploit its final results.

Contact :

Dipl.-Ing. Stephan Volmer (volmer@igd.fhg.de)

Relevant links :

Official FORMULA Web Site

Fast Multiresolution Image Querying

Fraunhofer IGD