An efficient self-collision handling between cloth surfaces based on spherical cluster technique

Simulating cloth model with realistic collision handling is a major issue in cloth simulation due to highly complex calculation. Numbers of techniques have been suggested by numerous researchers to look into the subject and produce efficient and robust collision handling for cloth simulation. In this paper, we introduce a new algorithm to reduce collision handling computation once cloth surfaces undergone wrinkle and twist process. The collision checking procedure between cloth surfaces can only be performed once a surface point contained by the predefined spherical clusters. Any surface point beyond the cluster region is ignored and removed from the collision calculation list. From the experiment conducted, as we compared with density-based clustering, series of brute force collision checking have been reduced due to a lesser amount of surfaces point within the spherical cluster are considered for collision checking. With this algorithm, we can guarantee that penetration among cloth surfaces would not occur and self-collision handling for cloth model is efficient and robust.