Cloth Fitting Based on Skeleton Driving and Implicit Skinning

In this paper, we propose a method of fitting a cloth model using skeleton driving and implicit skinning. Our method can automatically position a 3D garment onto a 3D human model without manual intervention of the user. As input, we use a cloth model and human model in different postures, we firstly generate a cloth skeleton by duplicating the skeleton of the human model. According to the skeletal information and the posture of the human model, we change the posture of the cloth using a set of implicit mesh sampled separately for each bone of the cloth skeleton. The deformed cloth is positioned onto the 3D human mesh by establishment of the correspondence via feature matching. The interpenetration between garment model and human model was solved by geometrical method. we use mass-spring system to simulate the final cloth model with realistic dressing result. The experimental results demonstrate that the proposed method presents an efficient way to redress various garments onto various human models.