Cartoon Animation and Morphing with Wavelet Curve Descriptor

  • Authors:
  • Gene C.-H. Chuang;C.-C. Jay Kuo

  • Affiliations:
  • Philips EBEI (Taiwan) Ltd., Taipei, Taiwan;Signal and Image Processing Institute and the Department of Electrical Engineering-Systems, University of Southern California, Los Angeles, California 90089-2564

  • Venue:
  • Multidimensional Systems and Signal Processing
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

A new framework for cartoon animation is proposed inthis work. By using the wavelet coefficients as the control points,one can manipulate curves so that shape metamorphosis occursalong resolutions as well as spatial locations. We model themotion of a cartoon character with the Lagrangian dynamic equationwhere the multiscale curve is driven by relevant forces. Thespatial and frequency localization property of the multiscalecurve model results in a sparse and diagonally dominant representationof the mass and stiffness matrices of the Lagrangian equationand hence the computation can be greatly simplified. To furthersimplify this model, we consider a model which consists of adecoupled system of ODEs. We then perform an experiment by capturingan image sequence with the locomotion of a walking dog, selectingsome key frames, and tracing the positions of limbs between thesekey frames. The best motion parameters are determined by usingthe least squares approximation. These extracted parameters arethen be used to animate cartoon characters of a similar typeof motion. The result shows the proposed curve descriptor withmultiscale structure and local control property is promisingin cartoon animation applications.