Interactive Exploration of Large Photo Libraries

Abstract

Visual information, in particular in form of images, is becoming increasingly important, and consequently efficient and effective tools for managing these rapidly growing collections are highly sought after. Interactive image database browsing systems provide an interesting alternative to retrieval-based approaches as they let the user explore an image dataset in an intuitive fashion. Based on content-based concepts, large image collections are visualised so that visually similar images are located close to each other in the visualisation space. Once displayed, the user can then interactively browse through the image collection. The main approaches to visualising and browsing image collections are mapping-based techniques, which are based on dimensionality reduction, and clustering-based methods, that group similar images together. In this paper, we highlight how these two approaches can be effectively combined to devise an intuitive image database navigation system that has low computational requirements, both offline and online, and organises images based on colour content on a spherical visualisation space while providing hierarchical access to large image datasets.