July 9, 2015 Tomohiro Fukuda * 1 Hirokazu Ban * 2 Katsuhito Yagi * 3 Junro Nishiie * 2 1 Development of High - definition Virtual Reality for Historical Architectural and Urban Digital Reconstruction - A Case Study of Azuchi Castle and Old Castle Town in 1581 - *1 Osaka University, Osaka , Japan *2 Omihachiman City Government, Shiga , Japan *3 Toppan Printing Co., Ltd., Tokyo , Japan
32
Embed
CAAD FUTURES 2015: Development of High-definition Virtual Reality for Historical Architectural and Urban Digital Reconstruction - A Case Study of Azuchi Castle and Old Castle Town
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
July 9, 2015
Tomohiro Fukuda*1 Hirokazu Ban*2
Katsuhito Yagi*3 Junro Nishiie*2
1
Development of High-definition Virtual Reality for Historical Architectural and Urban Digital Reconstruction- A Case Study of Azuchi Castle and Old Castle Town in 1581 -
*1 Osaka University, Osaka, Japan
*2 Omihachiman City Government, Shiga, Japan
*3 Toppan Printing Co., Ltd., Tokyo, Japan
Contents
1. Introduction
2. Development of High-definition Virtual Reality Application
3. Validation of Qualitative Method
4. Quantitative Investigation
5. Conclusion
2
3D Visualization in architectural and urban fields
Experience and understand 3D virtual space intuitively
Began with the pre-rendering technique to create 3DCG image and animation
NTSC (640*480 px), 30 frames/sec, 30 min. to generate a frame3
1. Introduction -1
”Osaka” (1983), Sasada lab., Osaka-U
Real-time rendering technique (VR. 1/30-1/60 sec. to generate a frame)
Gives users to interact with 3DVE: Walk-through, comparison of design alternatives,
dynamic simulation in real time
Become common in the use of PCs
Problems of VR for arch. and urban use
Number of defining objects is large; Objects from a small scale to a huge scale is handled at
the same time -> Difficult to render in real-time with high accuracy and realism
▶ Texture mapping instead of geometry, Culling
▶ Level of Detail (LOD) to change 3D models and textures to depend on the distance from VR
camera 4
1. Introduction -2
VR screen capture Shimonoseki prj. (2010)
Previous Studies about VR application
Future simulation at design stage
Digital archiving of existing and past architecture and townscapes
►Studies to construct a detailed realistic VR using recent advances in the
2. Development of High-definition Virtual Reality Application
3. Validation of Qualitative Method
4. Quantitative Investigation
5. Conclusion
21
3.1. Methodology
Subjects: Three VR creators
(30s male, 30s female and 20s female)
Subjects described the pros and cons of
the representative VR scene by referring to
the evaluation items
Full view: included all of Azuchi Castle, the
old castle town and terrain (Time zone: 3)
Keep tower view (Time zone: 3)
Castle town view: looked at the old castle
town (Time zone: 3)
Main road view: looked up at the keep tower
from the main road (Time zone: 4)
22
3. Validation of Qualitative Method
PC spec PC model HP Z820 Workstation
OS Windows7 Professional 64-bit
CPU Intel Xeon E5-2643 v23.5GHz
Main
memory
64GB
Graphics
board
NVIDIA Quadro K6000 (VRAM:
12GB
Display Resolution 1920x1080
PC specifications
Factor Element Evaluation items
3D
modelingShape
Outline
Surface
Detail
Extensiveness
Rendering
LightingLight source
Color shade
Shade and
shadow
Shade
Shadow
Material
Color
Pattern
Texture
DepthFocus
Haze
Scene Natural scene
Image qualityColor
Aliasing
Expressing
Image
synthesis
Color
Outline
Composition
Staffage
Staffage
Scale
Perspective
Garnish
Time
Motion
Effect
Output
Condition
Dimension
Effect
3.2. Result and discussion – Full View
Positive (N=8): Integrated set of VR
images was realized.
Negative (N=4): Green texture
rendered in Mt. Azuchi was
unnatural.
23
3. Validation of Qualitative Method
3.2. Result and discussion – Keep tower View
24
3. Validation of Qualitative Method
Positive (N=8): Azuchi Castle model
was represented in detail
Negative (N=8): A part of the
daytime scene showed inconsistency
between shade and shadow.
Shade and shadows should be
calculated integrally based on the
physical quantity of light and
material properties, but wide area
real-time rendering techniques, it is
hard to apply the global illumination
technique.
Then, the shade data by IBL and the
shadow data by shadow mapping
had to be defined separately and
integrated.
3.2. Result and discussion – Castle town View
25
3. Validation of Qualitative Method
Positive (N=4): Modeling quality of
the houses was high.
Negative (N=13): Liveliness of the
castle town could not be observed.
Vario-items such as people, small
animals and store curtains should
be placed in the old castle town to
add liveliness.
But, they were placed manually with
much labor and expense.
3.2. Result and discussion – Main road View
26
3. Validation of Qualitative Method
Positive (N=7): Roof tiles and stone
walls were expressed realistically.
Negative (N=3): Consistency of the
sky and lighting were insufficient.
Real photos were used as textures of
the sky and wall stones. But, buildings
were calculated using CG on the PC as it
no photos existed. The inadequate
consistency resulted from this reasons.
Contents
1. Introduction
2. Development of High-definition Virtual Reality Application
3. Validation of Qualitative Method
4. Quantitative Investigation
5. Conclusion
27
4.1. Methodology
Subjects: End-users (N=286)
Subjects viewed both short movie and manual modes, they
responded to the evaluation after the symposium held in March
and May 2014
Questionnaire items were answered using Likert five-point
agree/disagree scale
28
4. Quantitative Investigation
VR experience by end-users
4.2. Result and discussion
Developed VR: “Very good”
and “Good” of respondents
accounted for 88% of
responses.
High-definition VR experience
of the old Azuchi Castle and its
castle town: “Very interesting”
or “Interesting” of respondents
accounted for 89%.
The developed VR application
received a very positive
evaluation by end users.
29
4. Quantitative Investigation
About the developed VR application (N=286)
About High-definition VR experience (N=286)
Contents
1. Introduction
2. Development of High-definition Virtual Reality Application
3. Validation of Qualitative Method
4. Quantitative Investigation
5. Conclusion
30
5. Conclusion
To perform 3D-VR real-time rendering with high resolution, accuracy and realis, both LOD and representation techniques of natural objects and phenomena were developed, and the VR application of Azuchi Castle and its old town was constructed by integrating them.
Through the qualitative and quantitative evaluations, the developed VR obtained a certain number of positive judgments. Problems to be solved such as shade and shadow, and lively representation of the castle town were also clarified.
This VR system will be used for tourism, education and civic pride.
31New theater for VR Azuchi Castle (April, 2015)Azuchi Castle Remains
32
Acknowledgements
We thank all the participants for their generous assistance in
conducting Azuchi Castle VR project and experiment.