PaperAnalysis / May
Monday, May 24th
After finishing the basics of the volume renderer I plan on integrating it into OpenGL?. I've been reading some texts about the depth buffer in OpenGL? and I've integrated some classes that will act as a wrapper in an obect oriented system around the depth buffer. However I will need to read some more to get it actually working.
This week I will otherwise mostly be occupied with stuff connected to the Image Analysis course since it's finals next week and my birthday is comming up.
Tuesday, May 18th
I've been tweaking my own raytracing algorithm during friday and monday. I've done some improvements among others I've done backface removal as well as fixed a bug where no rotation of the volume caused failure in rendering.
As you can see by the image below the quality of shading is somewhat worse than with vtk. There also seems to be some skew in the layers which is strange. I don't know if that is correct but I don't believe it was there in the vtk images.
Thursday, May 13th
The last two days I've been working on my raytracer again and thus far I've been able to write the line drawing algorithm in 3d (an extension of bresenhams's in 2d). I've also written a converter from pbm to vol which is my own volume representation and I've done some preliminary renderings of the volume with the raytracer and initial testings show good results (as I expected the hitcalculations are speeded by the relative denseness of the volume).
Image of the day. A Raytraced image, it seems a bit wrong compared to the others, but still it's a start.
Tuesday, May 11th
The first image of the volume is produced (yay for me). It's nothing special really. I have yet to enable shading in vtk but it's something.
I cooked together abit of shading:
And also another image more suitable for printing.
Thursday, May 7th
Managed to get ahold of the image data I need to construct my volume here's a sample image:
This is in pretty low resolution but as you can see there is a lot of surface in the volume and not much "volume" which ironically makes it suitable for volume rendering and not surface rendering.
The resolution is: 2551 x 768 x 102 which is abit smaller than the original image.
Wednesday, May 6th
I Basically just installed vtk on my office machine today.
Wednesday, May 5th
Ok, I've been a bad updater. But the hitcalculations on the cube are basically done. What I've been working on (up until friday) is the actual linedrawing in the cube which is not as simple as I first thought. However on thursday afternoon Xavier Tizon had a presentation on tools used for scientific visualization and mentioned VTK. I had read about that toolbox before but had forgotten about it. I'm now working on understanding VTK and I have built some simple applications using it. My idea is to test rendering a papervolume using the vtk, if it seems to work I could use my more memory efficient implementation of the volume to represent it and thus save some memory.
<< Back To The Diary