No abstract

Current technology has made possible to scan volumes with ease and thus huge amount of data is generated. Devising algorithms to efficiently store and index this data is an open challenge. In this article we propose a theoretical approach to volume indexing by extending the idea of using fractal compression to index images, to the third dimension. The described technique uses a classification method that greatly improve the exponential complexity of the volume fractal coding. The compressed volume can then be used to be indexed in a digital library in a more efficient way. Then, content-based similarity queries can take full advantage of the self-similarity property of fractals. All these features will be present in an application that is currently under development.