[vtkusers] bit wise operations for vtkImageData
David Gobbi
david.gobbi at gmail.com
Fri Mar 25 15:48:34 EDT 2011
In python it is possible to access the VTK data arrays directly via
python's buffer interface, so that's what I use when I want to quickly
move pixel data back and forth from python to VTK. It allows me to
use a VTK array (including the one that the image uses to store its
pixels) as if it was a native python array.
I have a feeling that the only way you will be able to rapidly do VTK
pixel operations of any kind from Java, is if someone adds a similar
feature to the Java wrappers.
- David
On Fri, Mar 25, 2011 at 1:31 PM, Jonathan Morra <jonmorra at gmail.com> wrote:
> Your understanding is correct, however I cannot
> use GetScalarComponentAsDouble and SetScalarComponentAsDouble because they
> are way too slow in Java. Do you have any other suggestions for how to do
> this or will I have to write it myself in C++?
>
> On Fri, Mar 25, 2011 at 12:29 PM, David Gobbi <david.gobbi at gmail.com> wrote:
>>
>> Hi Jonathan,
>>
>> If I understand correctly, you have several binary images with exactly
>> the same dimensions. And you want to collapse them into a single
>> image, by storing a bitfield in the pixels of that image. So:
>>
>> To set bit "i" in pixel "xyz" (C++ code):
>> int a = int(image.GetScalarComponentAsDouble(x, y, z, 0));
>> a |= (1 << i);
>> image.SetScalarComponentAsDouble(x, y, z, a);
>>
>> To clear bit "i" in pixel "xyz":
>> int a = int(image.GetScalarComponentAsDouble(x, y, z, 0));
>> a &= ~(1 << i);
>> image.SetScalarComponentAsDouble(x, y, z, a);
>>
>> To test bit "i" in pixel "xyz":
>> int a = int(image.GetScalarComponentAsDouble(x, y, z, 0));
>> return ((a >> i) & 1);
>>
>> But I'm not sure if this is what you are trying to do. I didn't fully
>> understand your pseudocode.
>>
>> - David
>>
>>
>>
>> On Fri, Mar 25, 2011 at 1:08 PM, Jonathan Morra <jonmorra at gmail.com>
>> wrote:
>> > I'm in Java and storing a large number of binary vtkImageData's. I know
>> > this is inefficient, and am searching for a better way to store them.
>> > Right
>> > now I have a hash of the vtkImageData's and a get/put function.
>> > Basically I
>> > want to mimic this get/put but with better storage. I was thinking one
>> > way
>> > to do this is to use bitwise logical operations to store the information
>> > in
>> > the binary masks. For instance, if we have 2 binary images, then we
>> > could
>> > store that information in 1 vtkImageData using the following pseudo
>> > code.
>> > private static final int IMAGE1_BIT_CHANNEL = 0;
>> > private static final int IMAGE2_BIT_CHANNEL = 1;
>> > private vtkImageData storedImage;
>> > vtkImageData get(String image) {
>> > int channel = image eq "image1" ? IMAGE1_BIT_CHANNEL :
>> > IMAGE2_BIT_CHANNEL;
>> > vtkImageData return = new vtkImageData();
>> > foreach (pixel in storedImage)
>> > if (pixel at bit channel)
>> > return[pixel] = 1;
>> > else
>> > return[pixel] = 0;
>> > return return;
>> > }
>> > void put(String image, vtkImageData binaryImage) {
>> > int channel = image eq "image1" ? IMAGE1_BIT_CHANNEL :
>> > IMAGE2_BIT_CHANNEL;
>> > foreach (pixel in binaryImage)
>> > if (pixel)
>> > storedImage at bit in channel = 1;
>> > else
>> > storedImage at bit in channel = 0;
>> > }
>> > This could be extended easily for 8 channels for a char image for
>> > instance.
>> > This operation would have to be very fast though cause it is done often
>> > on
>> > the UI thread.
>> > 1. Is there a way to do this in VTK with Java?
>> > 2. Is this the best scheme for accomplishing my goal?
>> > 3. Is there a better scheme for doing this?
>> > Thanks.
>> > _______________________________________________
>> > 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 VTK FAQ at:
>> > http://www.vtk.org/Wiki/VTK_FAQ
>> >
>> > Follow this link to subscribe/unsubscribe:
>> > http://www.vtk.org/mailman/listinfo/vtkusers
>> >
>> >
>
>
More information about the vtkusers
mailing list