Automated Mosaics via Topology Inference

  • Authors:
  • Steve Hsu;Harpreet S. Sawhney;Rakesh Kumar

  • Affiliations:
  • -;-;-

  • Venue:
  • IEEE Computer Graphics and Applications
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

This article presents a complete approach for automated construction of mosaics from images and video, constituting a practical end-to-end system. Local alignment of spatially overlapping frames followed by global consistency provides spatial continuity, while compositing via multiresolution blending provides photometric continuity, so that the mosaic is free of seams that betray the underlying frames. When the camera traverses the scene with a 2D scan pattern, it is necessary to identify and align all spatially overlapping frames, including those not temporally contiguous. To reduce the amount of user interaction and to relax constraints on image capture, the novel topology inference algorithm will automatically determine a graph of spatial overlap relationships within a collection of frames, regardless of the scan pattern. We will highlight the generality of our approach by its application to planar mosaics, spherical mosaics, and geocoded terrestrial mosaics.