Image Database Navigation on a Hierarchical MDS Grid

Abstract

Due to the ever growing size of image collections with commercial image databases in excess of 1,000,000 images, efficient and effective ways of browsing and navigation through them as well as locating and searching desired images are in high demand. An interesting approach of providing a navigational tool for image databases has been the application of multidimensional scaling (MDS) where thumbnails of images are projected onto a 2-dimensional space in such a way that the original similarities between images are best preserved. Unfortunately, MDS is both computationally expensive and is only of limited use for large image sets as images are occluding each other while at the same time certain parts of the projection space are not utilised. In this paper we provide an MDS based image database navigation approach that does not suffer from these disadvantages. Based on an initial MDS calculation images are placed on a regular grid which avoids any overlapping effects. Large image datasets are handled through a clustering technique which allows browsing in a hierarchical manner.