Efficient query by image content for very large image databases

Abstract

Search techniques of image data bases are considered. Called query by image content (QBIC), these techniques are capable of searching very large (in the range of hundreds of thousands) sizes of image data bases in the fields of art, medicine, photo-agencies, etc. QBIC techniques serve as ″information filters″ and simply reduce the search for the user who will ultimately discard false calls. The authors cover the background and most important previous work, main algorithms used, point out important similarities and differences with typical pattern recognition systems, and touch upon possible applications. A prototype system is also described.