Prev: How to add interactivity in a dynamic model in SIMULINK?
Next: weighted average time series
From: Krishna C on 7 Mar 2010 12:56
Hi friends,

Can any body help me in sorting our the following problem. I'm struck with it and in urgent need of it...

Here is the micro-graph I want to analyze..

http://picasaweb.google.com/chaitanya.alur/Thesis#5445892412162022978

Each part in it is a grain surrounded by grain boundaries and the black portion is the damage. I want to find the number of branches (i.e. number of grains boundaries) each damaged portion is connected with. A grain boundary is a black line between two grains. During this process i did some transformations to the above image and got the following results. Check out the images, they may be helpful..

http://picasaweb.google.com/chaitanya.alur/Thesis#5445892412845832290

http://picasaweb.google.com/chaitanya.alur/Thesis#5445892410862607874
and
http://picasaweb.google.com/chaitanya.alur/Thesis#5445892414214258658

all images are binary..except the last one..

i tried to use.. bwmorph-->branchpoints, but didn't worked out well..

seems line detection mask will be helpful..but donno exactly how to implement it..

Kindly anybody help me in sorting this out..

Many thanks in advance.. Thank You.. :)
From: ImageAnalyst on 7 Mar 2010 15:59
The branchpoints option of bwmorph() should work. Line detection may
not - after all, you already have lines (curves) so you'd just be
finding the same thing you already have. You'd need to point out why
bwmorph doesn't work.

If anyone else wants to try to help, it might be helpful to get some
background:
http://www.mathworks.com/matlabcentral/newsreader/view_thread/264386#715428
 | 
Pages: 1
Prev: How to add interactivity in a dynamic model in SIMULINK?
Next: weighted average time series