Using Fractal Coding to Index Image Content for a Digital Library
Biao Cheng, Aidong Zhang, Raj S. Acharya, Cláudio H. Sibata · 1995
Efficiently storing and retrieving images in a digital library have recently become active topics. Due to the limitation of traditional keyword-based retrieval approach, content-based image retrieval becomes increasingly demanding in today's digital library applications. Using fractal image compression technique, we have observed that its self-transformation property to compress images can be utilized to automatically index images by content. In this paper, we investigate the effect of fractal image coding on content-based image indexing. The initial experiments conducted on biomedical images demonstrate that this approach not only greatly reduces image storage but also is effective on indexing images by content. We thus believe that fractal coding provides a promising indexing approach to digital libraries. 1 Introduction The need to store and retrieve image data in computer is growing rapidly in many fields, including business, manufacturing, education, Computer Aided Design (CAD), ...