Transcript Klaus Mueller - Stony Brook University

Volume Graphics
What’s in the cards…
The Panelists
Kwan-Liu Ma
The Panelists
Kwan-Liu Ma
Min Chen
The Panelists
Kwan-Liu Ma
Min Chen
Baoquan Chen
The Panelists
Kwan-Liu Ma
Min Chen
Baoquan Chen
Michael
Meissner
The Panelists
Kwan-Liu Ma
Min Chen
Baoquan Chen
Michael
Meissner
Klaus Mueller
The Good Cards
wide
acceptance
The Good Cards
wide
acceptance
available
data
The Good Cards
wide
acceptance
available
data
lots of
research
The Good Cards
wide
acceptance
available
data
lots of
research
speed
(GPU)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Ten Issues at VG99
•
•
•
•
•
•
•
•
•
•
Storage (polys vs. voxels)
Effects (reflections, refractions, shadows)
Radiosity – is it easier/better with voxels?
Potential for modeling
Can the “Visible Human” walk?
Volume test data (a teapot with actual tea)
Role of image processing and computer vision
A stack of images is a volume (modeling)
Availability of real-time volume rendering
Penetration of volume graphics into other
disciplines (think Siggraph…)
Issues: Reality Check
• Storage (polys vs. voxels)

storage (texture memory)
Issues: Reality Check
• Storage (polys vs. voxels)

storage (texture memory)
• Effects (reflections, refractions, shadows)

effects (illustrative volume rendering)
Issues: Reality Check
• Storage (polys vs. voxels)

storage (texture memory)
• Effects (reflections, refractions, shadows)

effects (illustrative volume rendering)
• Radiosity – is it easier/better with voxels?

simulation of amorphous phenomena
Issues: Reality Check
• Storage (polys vs. voxels)

storage (texture memory)
• Effects (reflections, refractions, shadows)

effects (illustrative volume rendering)
• Radiosity – is it easier/better with voxels?

simulation of amorphous phenomena
• Potential for modeling

deformation with haptics
Issues: Reality Check
• Storage (polys vs. voxels)

storage (texture memory)
• Effects (reflections, refractions, shadows)

effects (illustrative volume rendering)
• Radiosity – is it easier/better with voxels?

simulation of amorphous phenomena
• Potential for modeling

deformation with haptics
• Can the “Visible Human” walk?

Yes!
courtesy of D. Silver
Issues: Reality Check
• Volume test data (a teapot with actual tea)

still not much, submit to volvis.org
Issues: Reality Check
• Volume test data (a teapot with actual tea)

still not much, submit to volvis.org
• Role of image processing and computer vision

much better understanding of filters, etc.
Issues: Reality Check
• Volume test data (a teapot with actual tea)

still not much, submit to volvis.org
• Role of image processing and computer vision

much better understanding of filters, etc.
• A stack of images is a volume

video visualization
Issues: Reality Check
• Volume test data (a teapot with actual tea)

still not much, submit to volvis.org
• Role of image processing and computer vision

much better understanding of filters, etc.
• A stack of images is a volume

video visualization
• Availability of real-time volume rendering

GPUs !!!
Issues: Reality Check
• Volume test data (a teapot with actual tea)

still not much, submit to volvis.org
• Role of image processing and computer vision

much better understanding of filters, etc.
• A stack of images is a volume

video visualization
• Availability of real-time volume rendering

GPUs !!!
• Penetration of volume graphics into other
disciplines (think Siggraph…)

3D textures, subsurface scattering, virtual voyage
New Issues
• User interfaces

transfer functions are a pain
• Modeling tool

surface splatting vs. volume splatting
• Large datasets are still a problem

multi-variate, multi-valued ones, too
• Strides in segmentation are direly needed

need to get features from the scanned datasets
• Better understanding of perceptional issues

how can we best accentuate the features we find
Panelists… GO
Kwan-Liu Ma
Min Chen
Baoquan Chen
Michael
Meissner
Ten Issues for 2005
1. Proof of reliability and accuracy
2. Make interface more simple and less daunting
3. Work closely with other disciplines
4. Visual data mining and analysis
5. Effective visualization, not so much exploratory
6. Make more popular for target groups
7. Incorporation of cognition and perception
8. Usability
9. Make it taste like beer
10. Make it taste like a lobster
Ten Issues for 2005
1. Proof of reliability and accuracy
2. Make interface more simple (create information interfaces)
3. Work closely with other disciplines
4. Visual data mining / analysis / feature extraction (segmentation)
5. Effective / illustrative visualization, not so much exploratory
6. Make more popular for target groups
7. Incorporation of cognition and perception
8 User study / validation / common framework for this
9. Framework to integrate algorithms (VolumeShop Pro)
10. Global illumination