From nobody@cs.Buffalo.EDU Mon Jul 20 09:39 EDT 1998 From: nobody@cs.Buffalo.EDU Date: Mon, 20 Jul 1998 09:39:14 -0400 (EDT) To: techreps@cs.Buffalo.EDU Subject: techrep: POST request Content-Type: text Content-Length: 1877 Comments: hadar ContactPerson: gsesfah@cs.buffalo.edu Remote host: pollux.cs.buffalo.edu Remote ident: gsesfah ### Begin Citation ### Do not delete this line ### %R 98-07 %U 98-07.ps %A Sheikholeslami, Gholamhosein %A Wang, Wenjie %A Zhang, Aidong %T A Model of Image Representation and Indexing in Image Database Systems %D July 20, 1998 %I Department of Computer Science, SUNY Buffalo %K content-based image retrieval, image decomposition %X Image database systems need to be built to support effective and efficient access to image data on the basis of content. In this process, significant features must first be extracted from image data in their pixel format. These features must then be classified and indexed to assist efficient retrieval of image content. However, the issues central to automatic extraction and indexing of image content remain largely an open problem. In this paper, we investigate effective block-oriented image decomposition structures to be used as a fundamental data model for representing image content in large image databases. A hierarchical structure $S^g$-tree is introduced to decompose database images. Each decomposition of an image segment generates n subsegments of equal-size that are uniformly distributed within the image segment. Our analysis illustrates a special case of S^g-tree, nona-tree decomposition, when n equals 9, is an effective and efficient block-oriented decomposition structure to facilitate content-based image retrieval. Using S^g-tree structure to represent image content in an image database, flexible access to partial contents of each database image can be provided. Various types of content-based queries and efficient image retrieval can then be supported through novel indexing and searching approaches.