A physically-based muscle and skin model for facial animation

Coull, Alasdair D. (2006) A physically-based muscle and skin model for facial animation. PhD thesis, University of Glasgow.

Full text available as:
Download (20MB) | Preview
Printed Thesis Information: https://eleanor.lib.gla.ac.uk/record=b2298011


Facial animation is a popular area of research which has been around for over thirty years, but even with this long time scale, automatically creating realistic facial expressions is still an unsolved goal. This work furthers the state of the art in computer facial animation by introducing a new muscle and skin model and a method of easily transferring a full muscle and bone animation setup from one head mesh to another with very little user input.

The developed muscle model allows muscles of any shape to be accurately simulated, preserving volume during contraction and interacting with surrounding muscles and skin in a lifelike manner. The muscles can drive a rigid body model of a jaw, giving realistic physically-based movement to all areas of the face.

The skin model has multiple layers, mimicking the natural structure of skin and it connects onto the muscle model and is deformed realistically by the movements of the muscles and underlying bones. The skin smoothly transfers underlying movements into skin surface movements and propagates forces smoothly across the face.

Once a head model has been set up with muscles and bones, moving this muscle and bone set to another head is a simple matter using the developed techniques. The developed software employs principles from forensic reconstruction, using specific landmarks on the head to map the bone and muscles to the new head model and once the muscles and skull have been quickly transferred, they provide animation capabilities on the new mesh within minutes.

Item Type: Thesis (PhD)
Qualification Level: Doctoral
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Colleges/Schools: College of Science and Engineering > School of Computing Science
Supervisor's Name: Patterson, Dr. John and Siebert, Dr. Paul
Date of Award: 2006
Depositing User: Angi Shields
Unique ID: glathesis:2006-3450
Copyright: Copyright of this thesis is held by the author.
Date Deposited: 15 Jun 2012
Last Modified: 10 Dec 2012 14:07
URI: http://theses.gla.ac.uk/id/eprint/3450

Actions (login required)

View Item View Item


Downloads per month over past year