2003 Fiscal Year Final Research Report Summary
Real-Tune 3D Shape Reconstruction, Visualization, Editing, and Coding for 3D Video
| Project/Area Number |
13308017
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (A)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Intelligent informatics
|
|---|
| Research Institution | KYOTO UNIVERSITY |
|---|
Principal Investigator |
MATSUYAMA Takashi MATSUYAMA,Takashi, 情報学研究科, 教授 (10109035)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
TOUKAI Syougo Fukui University, Faculty of Engineering, Associate Professor, 工学部, 助教授 (50283627)
SUGIMOTO Akihiro National Institute of Genetics, Human-Machine Symbiosis Research, Associate Professor, 知能システム系, 助教授 (30314256)
WADA Toshikazu Wakayama University, Faculty of Systems Engineering, Professor, システム工学部, 教授 (00231035)
HABE Hitoshi KYOTO UNIVERSITY, Graduate School of Engineering, Research Associate, 工学研究科, 助手 (80346072)
KAWASHIMA Hiroaki KYOTO UNIVERSITY, Graduate School of Informatics, Research Associate, 情報学研究科, 助手 (40346101)
|
|---|
| Project Period (FY) |
2001 – 2003
|
| Keywords | 3D Video / Multi-Viewpoint Video / Real-Time 3D Shape Reconstruction / Deformable Mesh Model / Lighting Environment Sensing / Texture Mapping / Free Viewpoint Video / PC Cluster |
|---|
| Research Abstract |
We have obtained the following technical attainments to realize 3D video generation, visualization, editing, and coding.
1. PC cluster system (PC : 30, active camera : 25) to capture synchronized multi-view video of human action
2. Precise calibration method for quasi fixed-viewpoint pan-tilt-zoom camera
3. 3 base plane visual cone intersection algorithm and its parallel pipeline processing method for real-time dynamic 3D shape reconstruction using PC cluster ; with this system, we can reconstruct over 10 volumes per second of human action at 2cm X 2cm X 2cm spatial resolution.
4. cooperative camera work planning for zoom-up active video capturing of human body parts ; with this method, we will be able to increase the accuracy of 3D shape and surface texture significantly.
5. 3D deformable mesh model for reconstructing accurate 3D human shape and motion ; with this method, we can reconstruct even concave portions of 3D human body.
6. Viewpoint dependent high fidelity texture mapping algorithm for 3D video visualization ; with,this method, we can visualize 3D video of almost standard video quality.
7. Lighting environment sensing (i.e. estimation of multiple light sources) using multiple Lambertian spheres
8. Lighting environment sensing using skeleton cube
9. Interactive 3D video editing system using 3D video and omni-directional panoramic video
10. Coding method of omni-directional video based on unfolded regular polyhedron
11. Coding method of 3D video based on planar mapping of 3D surface shape
As for two coding methods listed above, we proposed them to MPEG for international standardization.
|
