ETD

Archivio digitale delle tesi discusse presso l'Università di Pisa

Tesi etd-09082012-125144


Tipo di tesi
Tesi di laurea specialistica
Autore
BALDACCI, ANDREA
URN
etd-09082012-125144
Titolo
Camera path optimization for film editing
Dipartimento
INGEGNERIA
Corso di studi
INGEGNERIA INFORMATICA
Relatori
relatore Dott. Ganovelli, Fabio
relatore Dott. Vecchio, Alessio
relatore Prof. Avvenuti, Marco
Parole chiave
  • film editing
  • Camera path optimization
  • Bézier path optimization
  • Bézier deformation
  • 3d film editing
  • quaternion fitting
  • quaternion interpolation
  • virtual camera path
Data inizio appello
04/10/2012
Consultabilità
Non consultabile
Data di rilascio
04/10/2052
Riassunto
In recent years, the rapid growth in the use of three dimensional graphics brought to the creation of a very large number of three dimensional models, sometimes made available in large databases. This evolution has pushed research towards fast 3d mesh search and retrieval and in building concise presentations visualizing salient views of the 3d mesh to the user. This thesis contributes to this research field with a novel framework to build a cinematographic video presentation of a tridimensional static scene. In our system, the user sketches the desired movements of the camera using common interfaces such as mouse and keyboard. The resulting shots are converted to a more compact and more easily editable representation based on a parametrization of the camera path using Bézier curves. Given this new representation, shots are edited to comply to cinematographic constraints and then the montage is performed, i.e. these shots are edited to assemble the final sequence. While in real filmmaking, the shots can essentially only be cut, we allow more complex editing, i.e. changing the trajectory and the camera orientation. We mathematically define the montage as a constrained optimization problem where the unknowns are the modifications of the input shots, and the constraints are the total running time and the stylistic choices, and where the objective function is the difference of the optical flow from the original input movements.
File