Computer Laboratory

Projects

Cross dissolve without cross fade:
preserving contrast, color and salience in image compositing

M. Grundland, R. Vohra, G. P. Williams, and N. A. Dodgson
Computer Graphics Forum 25 (3), pp. 577–586, (Proc. Eurographics 2006), ISSN 0167-7055

Teaser image

Linear interpolation is the standard image blending method used in image compositing. By averaging in the dynamic range, it reduces contrast and visibly degrades the quality of composite imagery. We demonstrate how to correct linear interpolation to resolve this longstanding problem. To provide visually meaningful, high level control over the compositing process, we introduce three novel image blending operators that are designed to preserve key visual characteristics of their inputs. Our contrast preserving method applies a linear color mapping to recover the contrast lost due to linear interpolation. Our salience preserving method retains the most informative regions of the input images by balancing their relative opacity with their relative saliency. Our color preserving method extends homomorphic image processing by establishing an isomorphism between the image colors and the real numbers, allowing any mathematical operation defined on real numbers to be applied to colors without losing its algebraic properties or mapping colors out of gamut. These approaches to image blending have artistic uses in image editing and video production as well as technical applications such as image morphing and mipmapping.

Paper:

Supplementary documents:

Supplementary videos:

  • Cross dissolve (8.8 MB AVI DivX)
    • Vase and lighthouse
      • Left: Linear cross dissolve
      • Right: Salience pyramid cross dissolve
    • Vineyard and city
      • Left: Linear cross dissolve
      • Right: Contrast pyramid cross dissolve
    • Gecko and mountain
      • Left: Linear cross dissolve
      • Right: Contrast pyramid cross dissolve
  • Image morphing (1.1 MB AVI DivX)
    • Parrot
      • Left: Linear morphing
      • Right: Salience pyramid morphing
  • Image filtering (2.2 MB AVI DivX)
    • Bee
      • Top left: Original image
      • Top right: Linear Gaussian smoothing
      • Bottom left: Color preserving Gaussian smoothing
      • Bottom right: Contrast preserving Gaussian smoothing
    • Flowers
      • Top left: Original image
      • Top right: Linear Gaussian smoothing
      • Bottom left: Color preserving Gaussian smoothing
      • Bottom right: Contrast preserving Gaussian smoothing

Contact:

Acknowledgements:

  • We wish to thank Malcolm Sabin, Piotr Goldstein, and Bruce Gooch for their insightful suggestions. We are grateful to Javier Portilla and Eero Simoncelli for making freely available their MatLab implementation of multiresolution pyramids. Mark Grundland wishes to thank his teachers, especially Prakash Panangaden, Godfried Toussaint, and Luc Devroye, for the beautiful puzzles, the answered questions, and the inspiration to always ask one more. Mark Grundland's graduate studies at the University of Cambridge were made possible by the generous support of the Natural Sciences and Engineering Research Council, le Fonds Quebecois de la Recherche sur la Nature et les Technologies, the Celanese Canada Internationalist Fellowship, the British Council, the Overseas Research Student Award Scheme, and the Cambridge Commonwealth Trust.