1 Traditional Animation Method
Traditional animation embraces many methods including cel animation, stop-motion animation, claymation (a.k.a clay animation) and further techniques based 3D CGI and motion capture. In the following we discuss about 3D CGI with an example of techniques used for face animation, and motion capture. 2D animation techniques and stop-motion information can be found in (El Rhalibi, A. and Shen, Y. (2004)).
2 Animation with 3D CGI
3D Computer-generated imagery (CGI) transformed animation. The first film completed entirely in 3D CGI was Toy Story (1995), produced by Pixar. The process of 3D CGI is very repetitive and similar in that sense to classical 2D animation and follows many of the same principles. 3D CGI use very sophisticated software applications to create and manipulate very complex and realistic 3D objects, characters and environments. The software also provides functionalities for the animation of 3D scene components.
The major difference between 3D CGI animation and traditional animation is in 3D CGI the images are replaced by 3D models which are digitally modelled using 3D modelling and animation tools. The 3D models are animated roughly like stop motion animation but in 3D CGI there is no physical object. The objects in the 3D world are all virtual objects. The CAA (Computer Aided Animation) widely used in 2D animation is a mix of both categories of animation. The traditional animation is done with the aid of computer to speed up the process of keyframe, cel, and interpolation. Some software languages and platforms, such as Flash which is a multimedia platform, provide these functionalities for both 2D and 3D.
3D CGI can be used for many kind of animation, including in combination with standard such as MPEG4. The following section reviews the use of 3D CGI for Facial Modelling and Animation.
3 Facial Modelling and Animation
We provide an overview into the current research conducted in regards to both facial modelling and facial animation techniques (N. Ersotelos and F. Dong, 2007).
Modelling a 3D virtual character can be cumbersome and time consuming process. Typically a character is modelled as a polygon mesh; whilst alternative techniques exist such as multi-layer modelling which include underlying structures such as the muscle and skeletal formations, these techniques are still considered too computationally expensive, particularly in the context of real-time applications. There are several approaches to constructing a modelled character: Standard Geometric Face Modelling, an approach typically used in animated feature films and games; Generic Model Individualization (GMI), where a specific model is constructed through the deformation of an existing model, such as those constructed from video streams and 3D face scanners; and Example-based Face Modelling (EFM), where a face is constructed with the desired facial features through the linear combinations of an existing face model collection, an example of this is the morphable facial modelling technique. In relation to facial animation, there are three prominent classifications for the techniques employed: A Simulation-based approach, in which synthetic facial movements are generated through the mimicking of facial muscle contractions, whose application is prominent in the field of medical-based visualization.