出版社:The Institute of Image Information and Television Engineers
摘要:We present a 3-D modeling system that uses hand gestures for quickly and easily designing freeform models. The user illustrates a 3-D freeform in the air and the system automatically constructs plausible 3-D polygonal surfaces. A stereo camera captures the users' hands. We obtain a set of a vast number of points in a 3-D space that presents the freeform. The points are organized using voxel representation to remove redundancy of the points. The organized points are rendered using the Marching cubes method. Our informal study on first-time users showed that a first-time user typically masters the operations within a minute and can construct interesting 3-D models within seconds.