n**i
发帖数: 39 | | f**n
发帖数: 155 | 2 They are very useful, especially if you are doing medical image analysis.
【在 n**i 的大作中提到】
: 大家觉得好用么
| o******r
发帖数: 259 | 3 For ITK, it has many stuff in image processing. However, I just
find its level set in 2D only, not 3D. And I use matlab to
deal with various image filtering, format conversion. In sum,
I read its doc carefully and run some samples, but I still stick to my
own way.
For Vtk, as I use matlab to do the visualization, I don't bother to
utilize those visulizing functions.
ITK and VTK have many functions, but I just find some replacement are easier
to accomplish my task.
【在 f**n 的大作中提到】
: They are very useful, especially if you are doing medical image analysis.
| l**p
发帖数: 474 | 4 are you doing volme data processing?
【在 o******r 的大作中提到】
: For ITK, it has many stuff in image processing. However, I just
: find its level set in 2D only, not 3D. And I use matlab to
: deal with various image filtering, format conversion. In sum,
: I read its doc carefully and run some samples, but I still stick to my
: own way.
: For Vtk, as I use matlab to do the visualization, I don't bother to
: utilize those visulizing functions.
: ITK and VTK have many functions, but I just find some replacement are easier
: to accomplish my task.
| o******r
发帖数: 259 | 5 Not exactly.
I use surface models for shape analysis, image processing.
Most of the time I deal with triangle meshes.
To visulize the 3D image segmentation result, I sometimes process
volume data like the marching cubes do. BTW, vtk has the MC code
and co-founder of vtk, Lorensen, invented MC.
【在 l**p 的大作中提到】
: are you doing volme data processing?
| l**p
发帖数: 474 | 6 How do you segment the 3D image? some new methods?
【在 o******r 的大作中提到】
: Not exactly.
: I use surface models for shape analysis, image processing.
: Most of the time I deal with triangle meshes.
: To visulize the 3D image segmentation result, I sometimes process
: volume data like the marching cubes do. BTW, vtk has the MC code
: and co-founder of vtk, Lorensen, invented MC.
| o******r
发帖数: 259 | 7 So far, everybody use level set. Some combines it with EM framework,
which is Expectation-Maximization.
【在 l**p 的大作中提到】
: How do you segment the 3D image? some new methods?
| l**p
发帖数: 474 | 8 I am also interested in the level set....can you give a introduction on this topic?
【在 o******r 的大作中提到】
: So far, everybody use level set. Some combines it with EM framework,
: which is Expectation-Maximization.
| f**n
发帖数: 155 | 9 What do you think of the future of level set?
It seems that most people work on modifying the speed function. No much more w
ork can be done in this area.
【在 o******r 的大作中提到】
: So far, everybody use level set. Some combines it with EM framework,
: which is Expectation-Maximization.
| o******r
发帖数: 259 | 10 Level set is so popular that ppl exploit every area possible.
Many publications on its maths stuff are in in comp. physics,
which I don't have access to.
My ability is quite limited to applications, which still contain a lot
of opportunities.
I see publications on better image segmentation on specific medical
structures, like vessel, surface smoothing and reconstruction.
I'm not qualified to talk about its future.
It's Sethian and Malladi's job. :)
【在 f**n 的大作中提到】
: What do you think of the future of level set?
: It seems that most people work on modifying the speed function. No much more w
: ork can be done in this area.
| o******r
发帖数: 259 | 11 Buddy, tons of references are available.
Just give you a list.
http://math.berkeley.edu/~sethian/Publications/publications.html
【在 l**p 的大作中提到】
: I am also interested in the level set....can you give a introduction on this topic?
| l**p
发帖数: 474 | 12 我是说你能不能总结一下讲一讲心得。
【在 o******r 的大作中提到】
: Buddy, tons of references are available.
: Just give you a list.
: http://math.berkeley.edu/~sethian/Publications/publications.html
| f**n
发帖数: 155 | 13 Good points!
I noticed that many good papers on level sets published on the
J. Comp. Physics. I think I need to pay more attention to it. ;)
【在 o******r 的大作中提到】
: Level set is so popular that ppl exploit every area possible.
: Many publications on its maths stuff are in in comp. physics,
: which I don't have access to.
: My ability is quite limited to applications, which still contain a lot
: of opportunities.
: I see publications on better image segmentation on specific medical
: structures, like vessel, surface smoothing and reconstruction.
: I'm not qualified to talk about its future.
: It's Sethian and Malladi's job. :)
|
|
