2004 Fiscal Year Final Research Report Summary
Transdimensional Processing of Captured Data Computation
| Project/Area Number |
14380137
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
計算機科学
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
YAMAGUCHI Yasushi The University of Tokyo, Interfaculty Initiative in Information Studies, Professor, 大学院・情報学環, 教授 (80210376)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TERASAWA Mikio Nihon University, College of Economics, Professor, 経済学部, 教授 (90267005)
TAKAHASHI Shigeo The University of Tokyo, Graduate School of Arts and Sciences, Associate Professor, 大学院・総合文化研究科, 助教授 (40292619)
|
|---|
| Project Period (FY) |
2002 – 2004
|
| Keywords | Image Data / Shape Data / Transdimensional Processing / View Morphing / Painterly Morphing / Volumetric Data / Visual Cryptography |
|---|
| Research Abstract |
It has become much popular and easy to use real-world data such as three dimensional data as well as image data for computation and communication. In fact, real world data is rapidly increasing like range image, volumetric data, digital photo and video, etc. This kind of real world data is generally huge so that the computational and communication costs to handle such data is quite expensive. This study aimed at developing the efficient techniques for handling real-world data based on the concept of 'transdinesional processing'. The three targets are 'interpolation of 2D image', 'analysis of 3D data', and 'analysis of color'.
Concerning the interpolation of 2D image, we made two achievements, namely 'painterly morphing using brush stroke model' and 'real-time view-morphing based on epipolar geometry.' The painterly morphing takes two images as input and generates a morphing sequence whose frames are all rendered in painterly style. The real-time view-morphing also takes two pictures of
… More
the same scene taken from the different view positions and synthesizes images from the in between viewpoints. This technique can generate images in real-time with texture mapping capability.
This study also succeeded in developing 'surperspective projection of landscapes for geographical guide-map generation' and 'topological volume skeletonization of scalar field' in the sense of 3D data analysis. The former study can automatically generates a guide-map by subdividing the terrain model along ridge lines and ravine lines. The latter one provides us a new measure for analyzing the scalar field with volume skeletonization. It is applicable to design a transfer function for volume rendering.
The study on color analysis achieved 'visual cryptography for natural images' and 'texture screening method for pencil drawing effects.' The visual cryptography is a kind of cryptography which can be decoded only with the human visual system. For instance, a hidden image appears if two OHP sheets with random dots are stacked together. The study developed a new technique for handling a natural image with high image quality. Less
|
Research Products
(15 results)
