Due to the multiplicity of the implementation, virtual garment simulation requires the
capacity of dressing a sewn garment onto various posed human models in a real time. In this paper, a
three-step method has been proposed to accelerate the dress-up procedure. The virtual garment is first
dressed onto a human model in a standing pose to record a distance map between the garment shell and
the human model. And then the distance map is employed to generate a coarse match when the pose of
the human model has been changed. Finally, a fine-tuned 3D configuration without surface penetration
is generated after a draping/relaxation scheme. Experimental results suggest this method is a useful and
fast treatment for dressing sewn virtual garment on various posed human models.
