Narrow Results

Image retrieval plays an important role in a broad spectrum of applications. Contentbased retrieval (CBR) is one of the popular choices in many biomedical and industrial applications. Discrete image transforms have been widely studied and suggested for many image retrieval applications. The Discrete Wavelet Transform (DWT) is one of the most popular transforms recently applied to many image processing applications. The Daubechies wavelet can be used to form the basis for extracting features in retrieving images based on the description of a particular object within the scene. This wavelet is widely used for image compression. In this paper we highlight the common features between compression and retrieval. Several examples are used to test the DWT retrieval system. A comparison between DWT and Discrete Cosine Transform (DCT) is also made. The retrieval system using DWT requires preprocessing and normalization of images, which might slow down the retrieval process. The accuracy of the retrieval using DWT has been significantly improved by incorporating efficient K-Neighbor Nearest Distance (KNND) measure in our system.

We propose a new shape-based, query-by-example, image database retrieval method that is able to match a query image to one of the images in the database, based on a whole or partial match. The proposed method has two key components: the architecture of the retrieval and the features used. Both play a role in the overall retrieval efficacy. The proposed architecture is based on the analysis of connected components and holes in the query and database images. The features we propose to use are geometric in nature, and are invariant to translation, rotation and scale. Each of the suggested three features is not new per se, but combining them to produce a compact and efficient feature vector is. We use hand-sketched, rotated and scaled query images to test the proposed method using a database of 500 logo images. We compare the performance of the suggested features with the performance of the moment invariants (a set of commonly-used shape features). The suggested features match the moment invariants in rotated and scaled queries and consistently surpass them in hand-sketched queries. Moreover, results clearly show that the proposed architecture significantly increases the performance of the two feature sets.

In this paper, a novel similar image retrieval scheme based on wavelet transformation will be presented. Our scheme is built upon a block-based query system. Our new scheme employs the wavelet transformation technique to transform each block in the spatial domain to the wavelet domain. Then, from each transformed block, the mean value and the edge types are extracted. These extracted features are then used to compute the similarity between a query image and the images in the database. In order to increase the similarity in the query result, the current block can be further divided into many sub-blocks, and then features can be extracted from these sub-blocks. Finally, the query result will be a set of ranked images in the database with respect to the query. According to our experiment, the proposed scheme can obtain satisfactory results.

A conceptual framework for image information systems is presented. Current research topics are surveyed, and application examples presented, followed by a discussion on the design issues for the next generation of image information systems. It is our view that the next generation of active image information systems should be designed based upon the notions of generalized icons and active indexes, resulting in smart images.