We propose a new polygonal approximation method for soft objects. While conventional polygonization methods decompose space into small-sized cells and compute many pieces of polygons for the cells, this new method polygonizes a soft object by smoothing an initial polygonal approximation using subdivision surface schemes and an exact computation scheme. The initial polygonal approximation is generated by the union of the polygonal approximations of the components that constitute the soft object. Our method has many advantages over conventional methods: faster computation time, regular-sized polygons, efficient robust results, and multiresolutional representation. These features will lead to an interactive modeling environment using soft objects as well as an efficient polygonal representation that is proper to be used in virtual reality and distributed environments.
Bibliographical noteFunding Information:
This research was supported in part by KOSEF (Korea Science and Engineering Foundation) under Grant No. NK9812801 and KRF (Korea Research Foundation) under Grant No. NH9733301.
All Science Journal Classification (ASJC) codes
- Signal Processing
- Human-Computer Interaction
- Computer Vision and Pattern Recognition
- Computer Graphics and Computer-Aided Design