THESIS
2012
x, 54 p. : ill. (chiefly col.) ; 30 cm
Abstract
Biomechanical human body modeling has a wide range of applications. However, there are no existing methods to construct accurate parametric human models. There are various factors that make this problem difficult, such as the difficulty to model free form shapes, large shape variations, poor models for soft tissues, etc....[
Read more ]
Biomechanical human body modeling has a wide range of applications. However, there are no existing methods to construct accurate parametric human models. There are various factors that make this problem difficult, such as the difficulty to model free form shapes, large shape variations, poor models for soft tissues, etc.
In this thesis, we introduce a new method to parametrically define shapes of skeleton components (bones). Our novel approach has several advantages: we require only a few sample bones of a class to construct a fairly robust template model; our approach does not rely on accurate identification of specific landmarks or shape features; finally, our algorithms are robust and efficient. The main tool we use is free form deformation (FFD), which we use to modify the shape of a template mesh by dislocating its affiliated lattice, which itself is automatically derived by mesh simplification. This template bone is automatically deformed to match the shape of different samples, given just a few parameter values as input. We also evaluate the accuracy of the model and analyze the performance of our approach under different settings.
Post a Comment