1998 Fiscal Year Final Research Report Summary
Modeling method for existing objects taking their characteristics in consideration
| Project/Area Number |
09680346
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
計算機科学
|
|---|
| Research Institution | Keio University |
|---|
Principal Investigator |
OHNO Yoshio Keio University, Faculty of Science and Technology, Professor, 理工学部, 教授 (20051865)
|
|---|
| Project Period (FY) |
1997 – 1998
|
| Keywords | geometric modeling / computer graphics / symmetry / rotational symmetry / mirror symmetry / wrinkly surface |
|---|
| Research Abstract |
We developed algorithms for geometric modeling of existing objects taking the characteristics of their shapes in consideration. The results are summarized below:
Detection of rotational symmetry : Location of center of gravity G is estimated from 3D data points, and converts locations of the data points into cylindrical coordinates using aribtrary line passing through G as the axis. The axis of rotational symmetry can be decided from the periodicity of the profile of projected data point set.
Geometric modeling using the rotational symmetry : Surface parameter is assigned to each data point, and the approximating surface patches are decided taking the axis of rotational symmetry in consideration.
Detection of mirror symmetry : Location of center of gravity G is estimated from 3D data points, and aribitray plane P that passes through G is chosen. Then the symmetry of the cross section S between the profile of data point set and P is decided by rotationg a line L around G on S and by detecting the local maxima and the local minima of the length of intersection between L and S.
Development of modeling methods based on the shape caracteristics : A class of surfaces that have both properties of fractal surfaces and spline surfaces are developed. We call these surfaces wrinkly surfaces. By using wrinkly surfaces, shapes with various levels of randomness and with various characteristics such as symmetry can be readily modeled.
Finally, we generalized the wrinkly surfaces and improved their applicability to wide range of shapes.
|
