Cost-effective storage and retrieval of images via the web
The project creates an innovative system that provides a solution to the problem of automatically storing and retrieving images from large image databases dispersed over the Internet. The business objectives of COBWEB are to reduce the time and costs to store a new image in an existing image database by at least 75%, dramatically reducing the need for textual annotation of images. Furthermore, it aims to reduce the time and costs to retrieve an image from an existing image database by at least 50%, by dramatically increasing the precision of the retrieval process. For that purpose, the project focused on the development of an innovative IT platform, offering advanced features for image analysis, developing effective algorithms for the automatic indexing of the visual features of imagery (e.g. colour, texture), that can be used as a means for retrieving similar images. The project also aims at the creation of a simple, useful and effective user interface, hiding the system's internal complexity, helping the user to perform queries and retrieve images in the most “natural” way. Therefore, the user interface provides retrieval tools, designed to match the human way of thinking. Finally, keeping in mind that the client's location may be remote, compared to the database's location, there will be support for cost-effective remote searches and access via the internet. Useful links: http://www.ImageQuery.com Other useful sites: Content-based Image Retrieval System, http://www.qbic.almaden.ibm.com/
