05-20-2010 03:04 AM
Yes I've tried that, if I do a normal threshold the little vains will disapeare,
so I added a Sobel and Prewit filter, than I could see the vains but than
there is extra noise.
In the attachment there is a colored picture and a green picture.
with the green pictures I work!
The green picture I load in the program, than it turns into a gray picture
and take out only the vains (as good as it can)
Hmm the coloured picture is to big, so I can only attach the green picture
if you want the coloured picture you could give me you email 🙂
05-20-2010 03:12 AM
05-20-2010 03:13 AM
05-20-2010 09:23 AM
The inverse Danielsson as you describe it would be interesting, but is not possible directly. The current version starts at the known outer edges and works its way in until it finds the center. There isn't an easy way to start at the center since it isn't known until after the Danielsson is complete.
If you AND your skeleton function with the Danielsson results, you should have all the width measurements along the center line of the vein. I don't know if that would do you any good.
What is your final goal with this measurement? Mean width? Median width? It is kind of difficult to get a vein network down to a single value. If you had the option of measuring manually, what exactly would you measure?
Bruce
05-21-2010 02:03 AM
My final goal is that I have a vain network with a thickness of one,
those lines I will color and the must show how thick the vain is.
this has to be accured to +/- 2 pixels.
Than I have to count all colors and say how many percent falls in
certain range, that I will have to save in an excel sheet.
Chantal
05-21-2010 04:21 AM
Try Skeleton M if at all it is going to help
You can also use the circle detection and use "Overlay oval" to overlay the diameter you want. You can also use different colors to overlay different range of diameters. You can also deduce the bounding box data from x,y,radius and overlay rectangle if you want.
05-24-2010 04:29 AM
05-26-2010 05:10 AM
Hello,
So far not real updates,
I'm trying to write an algorithem but that takes time 🙂
I think the danielson functies is the solution for this problem, but I will also use the circle detection.
When my program is ready I will updat again!
Chantal
05-26-2010 05:14 AM
07-01-2010 03:26 AM
Hello,
I'm almost fineshed with my program,
but there only one ting I wanted to add,
it's probably verry simpel, but I dont't know how to do it.
If the users starts the VI, I want the user to pick a photo,
but how do I make a VI wich atomaticly asks the user to pick a picture?
Thanks!
Chantal