Discussion:
[Paraview] How to do volume rendering of image data?
Steve Huntley
2010-07-15 08:37:09 UTC
Permalink
According to the Paraview FAQ, Paraview supports volume rendering of
image data. But despite searching the Web and buying the Paraview book,
I have not been able to find any detailed descriptions of this feature,
let alone directions for or examples of how to do it.

I have several image stacks in png format which represent slices of
3-dimensional objects. I can convert an image stack into a vtk file and
open the file in Paraview, but after that I'm at a loss. Is it possible
to do a volume rendering and end up with a photorealistic image of the
reconstructed 3D object?

After opening the vtk file created from the image stack and choosing
volume representation, all I see is a black block. The black is
evidently the background image area surrounding my 3D object. I would
like to make the black transparent and thus reveal the object, but it
seems to me that creating a custom color scale to control opacity would
result in color mapping and interpolation which would destroy much of
the original color in my image. Is there a way to declare a single
color transparent without affecting the other colors in the image? Is
this a valid way to pursue volume rendering of image data?

Up to now I have been using Osirix to reconstruct 3D objects from stacks
of slices and do photorealistic renderings of them. But Osirix has not
been able to handle my larger image sets without crashing, and I would
also like to be able to script the process, as well as take advantage of
Paraview's parallel processing capabilities.

Is it possible to accomplish something like Osirix's 3D reconstructions
in Paraview?


Thanks,
Steve Huntley
Guillaume.Duclaux
2010-07-15 08:51:36 UTC
Permalink
Hi Steve,

I don't know if what you describe can be achieved in Paraview, but actually you can do it easily with VolView [1] (another - bt not open-source - Kitware software).
Volume rendering is a standard technique for visualising X-Ray Computed Tomography data.
If you find out out to do this in Paraview, please share your knowledge!

Cheers

Gilly

[1] http://www.kitware.com/products/volview.html
Post by Steve Huntley
According to the Paraview FAQ, Paraview supports volume rendering of
image data. But despite searching the Web and buying the Paraview book,
I have not been able to find any detailed descriptions of this feature,
let alone directions for or examples of how to do it.
I have several image stacks in png format which represent slices of
3-dimensional objects. I can convert an image stack into a vtk file and
open the file in Paraview, but after that I'm at a loss. Is it possible
to do a volume rendering and end up with a photorealistic image of the
reconstructed 3D object?
After opening the vtk file created from the image stack and choosing
volume representation, all I see is a black block. The black is
evidently the background image area surrounding my 3D object. I would
like to make the black transparent and thus reveal the object, but it
seems to me that creating a custom color scale to control opacity would
result in color mapping and interpolation which would destroy much of
the original color in my image. Is there a way to declare a single
color transparent without affecting the other colors in the image? Is
this a valid way to pursue volume rendering of image data?
Up to now I have been using Osirix to reconstruct 3D objects from stacks
of slices and do photorealistic renderings of them. But Osirix has not
been able to handle my larger image sets without crashing, and I would
also like to be able to script the process, as well as take advantage of
Paraview's parallel processing capabilities.
Is it possible to accomplish something like Osirix's 3D reconstructions
in Paraview?
Thanks,
Steve Huntley
_______________________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html
Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView
http://www.paraview.org/mailman/listinfo/paraview
Moreland, Kenneth
2010-07-15 18:31:11 UTC
Permalink
Steve,

Your approach is correct. I don't know why you are seeing what you describe. How are you coloring your data?

-Ken


On 7/15/10 4:37 AM, "Steve Huntley" <stephen at xhuntley.net> wrote:

According to the Paraview FAQ, Paraview supports volume rendering of
image data. But despite searching the Web and buying the Paraview book,
I have not been able to find any detailed descriptions of this feature,
let alone directions for or examples of how to do it.

I have several image stacks in png format which represent slices of
3-dimensional objects. I can convert an image stack into a vtk file and
open the file in Paraview, but after that I'm at a loss. Is it possible
to do a volume rendering and end up with a photorealistic image of the
reconstructed 3D object?

After opening the vtk file created from the image stack and choosing
volume representation, all I see is a black block. The black is
evidently the background image area surrounding my 3D object. I would
like to make the black transparent and thus reveal the object, but it
seems to me that creating a custom color scale to control opacity would
result in color mapping and interpolation which would destroy much of
the original color in my image. Is there a way to declare a single
color transparent without affecting the other colors in the image? Is
this a valid way to pursue volume rendering of image data?

Up to now I have been using Osirix to reconstruct 3D objects from stacks
of slices and do photorealistic renderings of them. But Osirix has not
been able to handle my larger image sets without crashing, and I would
also like to be able to script the process, as well as take advantage of
Paraview's parallel processing capabilities.

Is it possible to accomplish something like Osirix's 3D reconstructions
in Paraview?


Thanks,
Steve Huntley
_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100715/a705c648/attachment-0001.htm>
Steve Huntley
2010-07-17 05:15:22 UTC
Permalink
Sure. Where should I send it?

In the meantime, is there a place on the web that gives examples of how
to do volume rendering of image data as specified in the FAQ?

--Steve H.
Hmm... That might be a gap in ParaView functionality. Do you have an
example data set you could send to the group?
-Ken
Thanks for responding. The thing is, I don't want to color my data at
all, I just want to have the original colors in the data file
displayed.
I uncheck the "map scalars" option, and this gives me the black
block. I
can use the slice filter to see that my properly colored data is
embedded in the block. But I see no way to force the black pixels
to be
transparent, so the data within is not visible.
Steve Huntley
2010-07-17 06:38:36 UTC
Permalink
Up to now I have been working with version 3.2. I just installed
version 3.8 to see if there were any differences in behavior.

In version 3.8 I'm not allowed to do volume rendering at all. When I
try to set the "Volume" representation option (which resulted in display
of a black block in 3.2), I now get an error message that says "Cannot
volume render since no point (or cell) data available."

Has the volume-rendering-of-image-data feature been excised from version
3.8?

--Steve H.
Hmm... That might be a gap in ParaView functionality. Do you have an
example data set you could send to the group?
-Ken
Thanks for responding. The thing is, I don't want to color my data at
all, I just want to have the original colors in the data file
displayed.
I uncheck the "map scalars" option, and this gives me the black
block. I
can use the slice filter to see that my properly colored data is
embedded in the block. But I see no way to force the black pixels
to be
transparent, so the data within is not visible.
Utkarsh Ayachit
2010-07-17 13:18:56 UTC
Permalink
Steve,

Try this:
* Create Wavelet source
* Hit Apply
* Then choose "Volume" from the representation combo.

That's an example of volume rendering of image data.

How let's see what's missing with your data so that it cannot be
volume rendered.
Look at the information tab. Do you see any single component
point-scalars? Mostly not. ParaView only supports volume rendering of
image data with single component point scalars.

Utkarsh
Up to now I have been working with version 3.2. ?I just installed version
3.8 to see if there were any differences in behavior.
In version 3.8 I'm not allowed to do volume rendering at all. ?When I try to
set the "Volume" representation option (which resulted in display of a black
block in 3.2), I now get an error message that says "Cannot volume render
since no point (or cell) data available."
Has the volume-rendering-of-image-data feature been excised from version
3.8?
--Steve H.
Hmm... That might be a gap in ParaView functionality. ?Do you have an
example data set you could send to the group?
-Ken
? ?Thanks for responding. The thing is, I don't want to color my data at
? ?all, I just want to have the original colors in the data file
? ?displayed.
? ?I uncheck the "map scalars" option, and this gives me the black
? ?block. I
? ?can use the slice filter to see that my properly colored data is
? ?embedded in the block. But I see no way to force the black pixels
? ?to be
? ?transparent, so the data within is not visible.
_______________________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://paraview.org/Wiki/ParaView
http://www.paraview.org/mailman/listinfo/paraview
Moreland, Kenneth
2010-07-17 13:43:52 UTC
Permalink
Nope. ParaView is just telling you that the current operation is not supported.

Probably the best course of action is to submit a feature request to ParaView's bug list (http://www.paraview.org/Bug) to which you can attach an example file. I would also recommend entering the feature in uservoice (http://paraview.uservoice.com/forums/11350-general). The ParaView developers use this site to determine the priority of suggested features.

If you have an example data set that is not too big, you can email it to the mailing list or directly to me and I can do these for you.

-Ken


On 7/17/10 2:38 AM, "Steve Huntley" <stephen at xhuntley.net> wrote:

Up to now I have been working with version 3.2. I just installed
version 3.8 to see if there were any differences in behavior.

In version 3.8 I'm not allowed to do volume rendering at all. When I
try to set the "Volume" representation option (which resulted in display
of a black block in 3.2), I now get an error message that says "Cannot
volume render since no point (or cell) data available."

Has the volume-rendering-of-image-data feature been excised from version
3.8?

--Steve H.
Hmm... That might be a gap in ParaView functionality. Do you have an
example data set you could send to the group?
-Ken
Thanks for responding. The thing is, I don't want to color my data at
all, I just want to have the original colors in the data file
displayed.
I uncheck the "map scalars" option, and this gives me the black
block. I
can use the slice filter to see that my properly colored data is
embedded in the block. But I see no way to force the black pixels
to be
transparent, so the data within is not visible.
_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100717/a6cc4967/attachment-0001.htm>
Steve Huntley
2010-07-18 08:43:25 UTC
Permalink
Attached is a small sample vtk file that illustrates the problem. It is
a rough representation of a human molar composed originally of 13 jpeg
image slices, processed into a raw file with ImageJ, then saved to vtk
by Paraview.

When I open the file in Paraview 3.2 I see a black block when I choose
volume representation. The information tab describes it as a uniform
rectilinear grid and shows number of points and cells, just like the
info tab of the sample wavelet object described by Utkarsh does.

When I open it in Paraview 3.8, the statistics entries in the
information tab are blank, and when I try to choose volume
representation I get the error mentioned below.

Utkarsh, I'm not sure what you mean by "single component point scalars."
Does this mean that color images with three scalar RGB values per grid
node can't be volume rendered? If so, does that mean that what I'm
currently attempting is impossible in Paraview? Or is there a way to
hack RGB color information into single-value scalars and render a color
3D image?
Nope. ParaView is just telling you that the current operation is not
supported.
Probably the best course of action is to submit a feature request to
ParaView?s bug list (http://www.paraview.org/Bug) to which you can
attach an example file. I would also recommend entering the feature in
uservoice (http://paraview.uservoice.com/forums/11350-general). The
ParaView developers use this site to determine the priority of
suggested features.
If you have an example data set that is not too big, you can email it
to the mailing list or directly to me and I can do these for you.
-Ken
Up to now I have been working with version 3.2. I just installed
version 3.8 to see if there were any differences in behavior.
In version 3.8 I'm not allowed to do volume rendering at all. When I
try to set the "Volume" representation option (which resulted in display
of a black block in 3.2), I now get an error message that says "Cannot
volume render since no point (or cell) data available."
-------------- next part --------------
A non-text attachment was scrubbed...
Name: molar.vtk.zip
Type: application/zip
Size: 31565 bytes
Desc: not available
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100718/ca637c48/attachment-0001.zip>
Steve Huntley
2010-07-19 08:34:04 UTC
Permalink
I tried a hackish solution to my problem by hand-crafting a version of
my vtk file which moved all the rgb data to a lookup table, and
converted the COLOR_SCALARS data set to a SCALARS set whose values are
indexes into the lookup table. I set the opacity of each color to 1.0,
except black which was set to 0.0. This in theory (as I currently
understand it) should have given me what I want.

The file opens in ParaView without error; however, it appears that the
lookup table is simply ignored, and the default color mapping is used
instead.

The custom vtk file is attached, in case anyone cares to take a look at
it and tell me if there's problem which causes ParaView to ignore the
lookup values, or if the behavior in this case is a bug.

Thanks,
Steve H.
Post by Steve Huntley
Attached is a small sample vtk file that illustrates the problem. It
is a rough representation of a human molar composed originally of 13
jpeg image slices, processed into a raw file with ImageJ, then saved
to vtk by Paraview.
When I open the file in Paraview 3.2 I see a black block when I choose
volume representation. The information tab describes it as a uniform
rectilinear grid and shows number of points and cells, just like the
info tab of the sample wavelet object described by Utkarsh does.
When I open it in Paraview 3.8, the statistics entries in the
information tab are blank, and when I try to choose volume
representation I get the error mentioned below.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: molar.look.vtk.zip
Type: application/zip
Size: 66392 bytes
Desc: not available
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100719/3fe4ed5e/attachment-0001.zip>
Moreland, Kenneth
2010-07-19 17:02:10 UTC
Permalink
The data you sent has RGB values but no alpha values that I can see. Volume rendering is generally not very useful if you do not have some representation of opacity. If you were to assume a homogeneous opacity, you would get a mostly black box with a blob of color in the middle if you looked hard enough. How exactly you expect the volume renderer to behave?

-Ken


On 7/18/10 2:43 AM, "Steve Huntley" <stephen at xhuntley.net> wrote:

Attached is a small sample vtk file that illustrates the problem. It is
a rough representation of a human molar composed originally of 13 jpeg
image slices, processed into a raw file with ImageJ, then saved to vtk
by Paraview.

When I open the file in Paraview 3.2 I see a black block when I choose
volume representation. The information tab describes it as a uniform
rectilinear grid and shows number of points and cells, just like the
info tab of the sample wavelet object described by Utkarsh does.

When I open it in Paraview 3.8, the statistics entries in the
information tab are blank, and when I try to choose volume
representation I get the error mentioned below.

Utkarsh, I'm not sure what you mean by "single component point scalars."
Does this mean that color images with three scalar RGB values per grid
node can't be volume rendered? If so, does that mean that what I'm
currently attempting is impossible in Paraview? Or is there a way to
hack RGB color information into single-value scalars and render a color
3D image?
Nope. ParaView is just telling you that the current operation is not
supported.
Probably the best course of action is to submit a feature request to
ParaView's bug list (http://www.paraview.org/Bug) to which you can
attach an example file. I would also recommend entering the feature in
uservoice (http://paraview.uservoice.com/forums/11350-general). The
ParaView developers use this site to determine the priority of
suggested features.
If you have an example data set that is not too big, you can email it
to the mailing list or directly to me and I can do these for you.
-Ken
Up to now I have been working with version 3.2. I just installed
version 3.8 to see if there were any differences in behavior.
In version 3.8 I'm not allowed to do volume rendering at all. When I
try to set the "Volume" representation option (which resulted in
display
of a black block in 3.2), I now get an error message that says "Cannot
volume render since no point (or cell) data available."
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100719/fbc87be0/attachment.htm>
Steve Huntley
2010-07-20 05:53:10 UTC
Permalink
The alpha values are in the lookup table. Each line in the lookup table
has four values: red, green, blue, opacity. That's my understanding of
a properly formatted lookup table in a vtk file. I used:
http://www.vtk.org/VTK/img/file-formats.pdf as a reference.

All opacity values are set to 1.0, except black which is set to 0.0.
But when the file is opened in ParaView and volume representation
chosen, it appears that not only the opacity values but the entire
lookup table is ignored.

Assuming homogeneous opacity is the furthest thing from what I want. I
want to find a way to control opacity without destroying the colors in
my image data file via mapping to the default color scale. Getting away
from homogeneous opacity is what I can't figure out how to do. I want
to see through the black background pixels to see the full color pixels
of my 3D object.

--Steve H.
Post by Moreland, Kenneth
The data you sent has RGB values but no alpha values that I can see.
Volume rendering is generally not very useful if you do not have some
representation of opacity. If you were to assume a homogeneous
opacity, you would get a mostly black box with a blob of color in the
middle if you looked hard enough. How exactly you expect the volume
renderer to behave?
-Ken
David E DeMarle
2010-07-20 16:02:30 UTC
Permalink
Unfortunately ParaView only has one-dimensional transfer functions
(and it doesn't handle opacity encoded in the alpha channel directly
all that well either). Please add this as a feature request.

More details for the curious:

We do one-dimensional transfer functions like so:

rgba' = f(value)

Where value is either a scalar (or a component or magnitude of a
vector to reduce it to a scalar)

You want a two dimensional transfer function. Specifically this very basic one:

rgb' = rgb, and a' = f(rgb)

Where f is a function of the magnitude of the rgb vector. We can't do
that because rgb and a can't be controlled independently within
ParaView.

The closest hack I could get to doing what you want is:

Apply this python filter to calculate a' = f(rgb):
pdi = self.GetInput().GetPointData()
colors = pdi.GetArray(0)
newcolors = vtk.vtkUnsignedCharArray()
newcolors.SetName("newcolors")
newcolors.SetNumberOfComponents(4)
for i in xrange(0,colors.GetNumberOfTuples()):
r,g,b = colors.GetTuple3(i)
o = sqrt(r*r+g*g+b*b)
newcolors.InsertNextTuple4(r,g,b,o)
pdo = self.GetOutput().GetPointData()
pdo.AddArray(newcolors)
pdo.SetScalars(newcolors)

(Note: producing o as an independent array lets you threshold to get
rid of the blank space in another way, but you still can't produce
rgb' = rgb in the volume rendering of that.)

Then use the slice representation and turn off map and interpolate
scalars to show the directly encoded alpha channel as opacity.

To fake volume rendering I applied the calculator filter then
repeatedly to produce multiple slices. When you look through (in the
right direction) you get a fake volume rendering effect. If you look
the wrong direction alpha blending doesn't work right.





David E DeMarle
Kitware, Inc.
R&D Engineer
28 Corporate Drive
Clifton Park, NY 12065-8662
Phone: 518-371-3971 x109
The alpha values are in the lookup table. ?Each line in the lookup table has
four values: red, green, blue, opacity. ?That's my understanding of a
http://www.vtk.org/VTK/img/file-formats.pdf as a reference.
All opacity values are set to 1.0, except black which is set to 0.0. ?But
when the file is opened in ParaView and volume representation chosen, it
appears that not only the opacity values but the entire lookup table is
ignored.
Assuming homogeneous opacity is the furthest thing from what I want. ?I want
to find a way to control opacity without destroying the colors in my image
data file via mapping to the default color scale. ?Getting away from
homogeneous opacity is what I can't figure out how to do. ?I want to see
through the black background pixels to see the full color pixels of my 3D
object.
--Steve H.
Post by Moreland, Kenneth
The data you sent has RGB values but no alpha values that I can see.
?Volume rendering is generally not very useful if you do not have some
representation of opacity. ?If you were to assume a homogeneous opacity, you
would get a mostly black box with a blob of color in the middle if you
looked hard enough. ?How exactly you expect the volume renderer to behave?
-Ken
_______________________________________________
Powered by www.kitware.com
Visit other Kitware open-source projects at
http://www.kitware.com/opensource/opensource.html
http://paraview.org/Wiki/ParaView
http://www.paraview.org/mailman/listinfo/paraview
Moreland, Kenneth
2010-07-20 17:19:39 UTC
Permalink
Uh, what lookup table? The data file you posted has no LOOKUP_TABLE entry. Even if it did, the VTK format specifies that the LOOKUP_TABLE is applied to a SCALARS entry whereas the molar.vtk file has data stored as COLOR_SCALARS. Even if the RGB values were listed under a SCALARS entry (which is an error in itself), the lookup table cannot be applied to a vector. VTK specifically defines a lookup table as something that converts scalars to colors. I'm not sure what the interface would look like for a lookup table that operated on vectors in either the VTK API or in a GUI.

As David DeMarle suggested in a separate email, it makes more sense to derive an alpha field and apply it to the color data. I'm still not sure ParaView supports directly volume rendering RGBA color values (the majority use case is to analyze general scalar fields), but perhaps it should.

-Ken


On 7/19/10 11:53 PM, "Steve Huntley" <stephen at xhuntley.net> wrote:

The alpha values are in the lookup table. Each line in the lookup table
has four values: red, green, blue, opacity. That's my understanding of
a properly formatted lookup table in a vtk file. I used:
http://www.vtk.org/VTK/img/file-formats.pdf as a reference.

All opacity values are set to 1.0, except black which is set to 0.0.
But when the file is opened in ParaView and volume representation
chosen, it appears that not only the opacity values but the entire
lookup table is ignored.

Assuming homogeneous opacity is the furthest thing from what I want. I
want to find a way to control opacity without destroying the colors in
my image data file via mapping to the default color scale. Getting away
from homogeneous opacity is what I can't figure out how to do. I want
to see through the black background pixels to see the full color pixels
of my 3D object.

--Steve H.
Post by Moreland, Kenneth
The data you sent has RGB values but no alpha values that I can see.
Volume rendering is generally not very useful if you do not have some
representation of opacity. If you were to assume a homogeneous
opacity, you would get a mostly black box with a blob of color in the
middle if you looked hard enough. How exactly you expect the volume
renderer to behave?
-Ken
_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview




**** Kenneth Moreland
*** Sandia National Laboratories
***********
*** *** *** email: kmorel at sandia.gov
** *** ** phone: (505) 844-8919
*** web: http://www.cs.unm.edu/~kmorel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100720/2a094b7b/attachment.htm>
Steve Huntley
2010-07-21 04:32:06 UTC
Permalink
I guess there was a communication problem. It seems you're still looking
at the first file I attached in an earlier message: molar.vtk. I created
a new file with the lookup table called molar.look.vtk and attached it
to the later message that discusses lookup tables and opacity.
Uh, what lookup table? The data file you posted has no LOOKUP_TABLE
entry. Even if it did, the VTK format specifies that the LOOKUP_TABLE
is applied to a SCALARS entry whereas the molar.vtk file has data
stored as COLOR_SCALARS. Even if the RGB values were listed under a
SCALARS entry (which is an error in itself), the lookup table cannot
be applied to a vector. VTK specifically defines a lookup table as
something that converts scalars to colors. I?m not sure what the
interface would look like for a lookup table that operated on vectors
in either the VTK API or in a GUI.
Moreland, Kenneth
2010-07-21 15:49:31 UTC
Permalink
Oh, I see now. I don't think this approach is going to work either. First of all, ParaView ignores the LOOKUP_TABLE defined in VTK file. It is an old feature that is rarely used anymore (the more recent VTK files don't support it), and it conflicts with ParaView's management of lookup tables. Even if ParaView accepted it, it would probably cause serious problems. The VTK classes are not prepared for lookup tables with this many values and are liable to fail in rather interesting ways.

-Ken


On 7/20/10 10:32 PM, "Steve Huntley" <stephen at xhuntley.net> wrote:

I guess there was a communication problem. It seems you're still looking
at the first file I attached in an earlier message: molar.vtk. I created
a new file with the lookup table called molar.look.vtk and attached it
to the later message that discusses lookup tables and opacity.
Uh, what lookup table? The data file you posted has no LOOKUP_TABLE
entry. Even if it did, the VTK format specifies that the LOOKUP_TABLE
is applied to a SCALARS entry whereas the molar.vtk file has data
stored as COLOR_SCALARS. Even if the RGB values were listed under a
SCALARS entry (which is an error in itself), the lookup table cannot
be applied to a vector. VTK specifically defines a lookup table as
something that converts scalars to colors. I'm not sure what the
interface would look like for a lookup table that operated on vectors
in either the VTK API or in a GUI.
_______________________________________________
Powered by www.kitware.com

Visit other Kitware open-source projects at http://www.kitware.com/opensource/opensource.html

Please keep messages on-topic and check the ParaView Wiki at: http://paraview.org/Wiki/ParaView

Follow this link to subscribe/unsubscribe:
http://www.paraview.org/mailman/listinfo/paraview




**** Kenneth Moreland
*** Sandia National Laboratories
***********
*** *** *** email: kmorel at sandia.gov
** *** ** phone: (505) 844-8919
*** web: http://www.cs.unm.edu/~kmorel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.paraview.org/pipermail/paraview/attachments/20100721/8ec3e93d/attachment.htm>
Loading...