On the Perceptual Organization of Image Databases Using Cognitive Discriminative Biplots
- PDF / 2,678,318 Bytes
- 15 Pages / 600.03 x 792 pts Page_size
- 66 Downloads / 138 Views
Research Article On the Perceptual Organization of Image Databases Using Cognitive Discriminative Biplots Christos Theoharatos,1 Nikolaos A. Laskaris,2 George Economou,1 and Spiros Fotopoulos1 1 Electronics
Laboratory, Department of Physics, University of Patras, 26500 Patras, Greece Intelligence and Information Analysis Laboratory, Department of Informatics, Aristotle University of Thessaloniki, 54124 Thessaloniki, Greece
2 Artificial
Received 14 December 2005; Revised 3 October 2006; Accepted 3 October 2006 Recommended by Maria Concetta Morrone A human-centered approach to image database organization is presented in this study. The management of a generic image database is pursued using a standard psychophysical experimental procedure followed by a well-suited data analysis methodology that is based on simple geometrical concepts. The end result is a cognitive discriminative biplot, which is a visualization of the intrinsic organization of the image database best reflecting the user’s perception. The discriminating power of the introduced cognitive biplot constitutes an appealing tool for image retrieval and a flexible interface for visual data mining tasks. These ideas were evaluated in two ways. First, the separability of semantically distinct image classes was measured according to their reduced representations on the biplot. Then, a nearest-neighbor retrieval scheme was run on the emerged low-dimensional terrain to measure the suitability of the biplot for performing content-based image retrieval (CBIR). The achieved organization performance when compared with the performance of a contemporary system was found superior. This promoted the further discussion of packing these ideas into a realizable algorithmic procedure for an efficient and effective personalized CBIR system. Copyright © 2007 Hindawi Publishing Corporation. All rights reserved.
1.
INTRODUCTION
The notion of image similarity has been exhaustively studied through the last decades in the field of computer vision. Its application includes recognition, classification, retrieval, and database organization, while its formulation is usually based on low-level attributes. Although such features (e.g., color, shape, texture, or combination of them) are considered to contribute to human judgment for image similarity, the idiosyncrasies of human perception are not fully considered at the algorithmic stage of feature extraction. The only exception is the recently developed research directions of relevance feedback [1, 2] and active learning procedures [3], in which the user is engaged to some iterative procedures aiming to alter the relative importance among the bunch of preselected features such that the modified machine-vision procedure matches his/her perception. Although significant progress has been reported since the early years [4], none of the existing methodologies are able to entirely encapsulate the semantic concepts necessary for expressing a high-level (i.e., cognitive) similarity between images. Information regarding visual perception is hi
Data Loading...
