Login
Username:

Password:

Remember me



Lost Password?

Register now!
Main Menu
H3D.org Feeds


Browsing this Thread:   1 Anonymous Users




« 1 (2) 3 »


Re: Volumerenderer and raw data
#11
Quite a regular
Quite a regular


See User information
It is not a secret, but our work is in the very early stages. We aim to analyse Optical Projection Tomography (OPT) data. The device can make single channel 3d images of translucent, generally biological, samples of size < 15 mm^3. Different filters can be used and in the future we aim to combine different channels into one image in order to show different characteristics of the speciment in one image. For shape analysis we plan to use point models, this is where the 3d pointing capabilities of haptics come in. Besides that we need a decent volume visualisation. Both haptics and volume visualisation are neatly combined in h3d. Hence our interest. I have changed the shader code in class Volumerenderer and now we get nice rgb output.
Johann

Posted on: 2007/1/10 9:07
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#12
Guru
Guru


See User information
Cool! Is it like confocal microscopy?

If you do changes to VHTK that you believe could be useful to others, feel free to send me a patch.

Posted on: 2007/1/10 9:34
KJ Lundin Palmerius
C Research
Linköping University, Sweden
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#13
Quite a regular
Quite a regular


See User information
OPT is not like confocal microscopy. The specimen is placed between a light source and a CCD camera. Hence the specimen has to be translucent in order for an image to be able to be captured by the camera behind the specimen. The specimen is then rotated in 400 discrete steps. Thus you end up with 400 images that each represent a projection of the specimen obtained from different angles around one axis. Those images are then fed into a standard tomography reconstruction algorithm to get your volume. The limiting factor is the maximum sample size resulting from the fact that the bigger the sample the more of the specimen is out of focus. The resolution is higher than conventional MRI and the machine is much cheaper than an MRI (Which for plant biologists is financially just not feasable). Furthermore you can image gene expression patterns...this is where it might get very, very interesting.
Johann

Posted on: 2007/1/10 11:47
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#14
Guru
Guru


See User information
That's really cool! Could you post a picture of a rendered data set?

Posted on: 2007/1/11 3:27
KJ Lundin Palmerius
C Research
Linköping University, Sweden
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#15
Quite a regular
Quite a regular


See User information
This is a h3d volume rendering of a three-channel flower bud image. This is a large image; I expect it gets scaled on display in the forum, let's see....Click to see original Image in a new window

Posted on: 2007/1/11 8:02
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#16
Quite a regular
Quite a regular


See User information
Right, this did not work the way I expected. Can I upload from my local filesystem or does the image need to have a real web url, i.e. be on a webserver already?

Posted on: 2007/1/11 8:08
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#17
Guru
Guru


See User information
Daniel just created an "other" category in the download section, for you to upload your image. Upload the image and then either reference it by name or create a link. I think that the uploaded image needs approval before being downloadable, so a reference by name is probably best.

Posted on: 2007/1/12 3:56
KJ Lundin Palmerius
C Research
Linköping University, Sweden
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#18
Quite a regular
Quite a regular


See User information
Hello,

I am having problems getting haptic feedback from my volume. My volume renderer uses the rgb data in form of a Image3DTexture. For the ScalarSurfaceFrictionMode I need a VHTKScalarDataNode (stipulated by the super-class of ScalarSurfaceFrictionMode). I have tried to derive this scalar data from my rgb volume by using MagnitudeExtractor. First I construct a VectorImageObjectVolume from my Image3DTexture in order to be able to pass it to the MagnitudeExtractor. I have also defined a transfer function, but I get no haptic feedback from the volume. I have also tried the same with class MagnitudeFilter, again without success. Programs run normally though, without crashes.
What am I doing wrong?
Johann

PS: Uploaded an image, got notified; but it is not in the download section. Must have given it the wrong download location.

Posted on: 2007/1/12 9:26
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#19
Webmaster
Webmaster


See User information
Re you image, now it works!

You may reach it from the header of your posting (Download section => other) or from here!

Thank you for showing this to the h3d community!

Regards
Johan
SenseGraphics

Posted on: 2007/1/12 10:12
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top


Re: Volumerenderer and raw data
#20
Guru
Guru


See User information
Really cool image! Great!

VHTK is not designed with RGB volumes or textures in mind, so I can't really tell you what is happening. All volume containers assume that the data is of scalar or vector contents. I think the magnitude extracted tries to convert Vec3f elements to its magnitude and in that case what's happening is that your RGB char data is not converted at all.

I think the easiest way to make it work is to convert the RGB texture offline and load it as a separate data set.

PS. Everything that is uploaded must be approved by the staff at SenseGraphics, just to stop spam and child pornography. That is why your upload was delayed.

Posted on: 2007/1/12 10:37
KJ Lundin Palmerius
C Research
Linköping University, Sweden
Twitter Facebook Google Plus Linkedin Del.icio.us Digg Reddit Mr. Wong Top




« 1 (2) 3 »



You can view topic.
You cannot start a new topic.
You cannot reply to posts.
You cannot edit your posts.
You cannot delete your posts.
You cannot add new polls.
You cannot vote in polls.
You cannot attach files to posts.
You cannot post without approval.
You cannot use topic type.
You cannot use HTML syntax.
You cannot use signature.
You cannot create pdf.
You cannot get print page.

[Advanced Search]





(C) 2012 SenseGraphics AB    ---    Powered by XOOPS