Molchanov,V., Linsen,L.
Abstract:
Deformation of geographical maps is a powerful tool used in geospatial data exploration systems. Virtual lenses help the user to retrieve local detail information without losing the global overview, while contiguous cartograms exploit map deformation for representing statistical data as areas of regions. In most applications, the map deformations should be smooth and computationally efficient. We propose a novel map deformation approach based on integral images computed for 2D density distributions. Depending on the prescribed density function, the resulting deformed map represents an approximation to a respective contiguous cartogram or may serve as a user-steered virtual lens. Our technique is suitable for use in highly interactive geospatial data exploration systems due to its algorithmic and computational efficiency. We test the proposed method on several artificial and real-world datasets and discuss directions for future work.