A cluster‐based approach for efficient content‐based image retrieval using a similarity‐preserving space transformation method

Abstract

The techniques of clustering and space transformation have been successfully used in the past to solve a number of pattern recognition problems. In this article, the authors propose a new approach to content-based image retrieval (CBIR) that uses (a) a newly proposed similarity-preserving space transformation method to transform the original low-level image space into a highlevel vector space that enables efficient query processing, and (b) a clustering scheme that further improves the efficiency of our retrieval system. This combination is unique and the resulting system provides synergistic advantages of using both clustering and space transformation. The proposed space transformation method is shown to preserve the order of the distances in the transformed feature space. This strategy makes this approach to retrieval generic as it can be applied to object types, other than images, and feature spaces more general than metric spaces. The CBIR approach uses the inexpensive "estimated" distance in the transformed space, as opposed to the computationally inefficient "real" distance in the original space, to retrieve the desired results for a given query image. The authors also provide a theoretical analysis of the complexity of their CBIR approach when used for color-based retrieval, which shows that it is computationally more efficient than other comparable approaches. An extensive set of experiments to test the efficiency and effectiveness of the proposed approach has been performed. The results show that the approach offers superior response time (improvement of 1-2 orders of magnitude compared to retrieval approaches that either use pruning techniques like indexing, clustering, etc., or space transformation, but not both) with sufficiently high retrieval accuracy.

Towards the semantic extraction of digital signatures for librarian image‐identification purposes

“…Finally this method may be characterized as an accurate method that produces a very good degree of complexity compared to other methods (see Shah et al, 2006). Furthermore, the method may be characterized as novel and helpful because of the algorithms used and the sophisticated schema based on this technique, which are in agreement with recent studies (Zachary et al, 2001; Shah et al).…”

“…In this section we will calculate the complexity of the procedure used in our method, and then we will compare the complexity of our method with that of a well-known, accurate philosophy method based on a similar philosophy (Shah et al, 2006). In our case, the query-image procedure is focused on the calculation of the complexity of the Hausdorff-distance extracted by two simple convex polygons that represent two images.…”

“…On the other hand, the method in Shah et al (2006) focuses on a new approach of the content‐based image retrieval (CBIR) problem, which is based on the combination of two methods: (a) the newly proposed similarity‐preserving space transformation method, which enables efficient query processing, and (b) a clustering scheme that further improves the efficiency of this retrieval system. The total optimal complexity of this method is calculated by the authors as O ( NL 2 ) time, where N is the number of images in the database and L is the number of low‐level features.…”

“…Finally, an additional aim of this article is to compare the proposed method with the studies by Zachary, Iyengar, and Barhen (2001) and Shah, Raghavan, Dhatric, and Zhao (2006), which are considered the most representative of the techniques of Web image retrieval. The comparison with these two studies is presented below.…”

Making colors worth more than a thousand words

“…Li et al [17] have performed architectonics building recognition using color, orientation, and spatial features of line segments. Raghavan et al [29] have designed a similarity-preserving space transformation method of low-level image space into a high-level vector space to improve retrieval. Some researchers have used bag of features for image categorization [10,13,21].…”