Crack details extraction

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Crack details extraction

anusuya pal
Hello,

I have an image (gray-scaled of the raw image) and I want to have the
detailed information about the cracks. Basically, I want to count all its
junctions, no of node points, branches and measure their average length. I
have found many threads discussed related to cracks in this forum. What I
come up is that the appropriate tool for my case will be *Analyze Skeleton
(2D/3D).*

But prior to use this tool, I need to binarise my image, and here the
problem comes. As I do that, the cracks do not exactly match with those in
the original image.

For the convenience, I have attached the image (n_varian.jpg) where I
applied variance filter just to get the cracks. And then tried to make it
binary, but it fails.

Are there any suggestions to get the details of the cracks or to solve this
issue?

Thanks
Anu

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

aaa.png (1M) Download Attachment
n_varian.jpg (674K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Crack details extraction

anusuya pal
Hello,

May I request anyone to look into this matter?
Unless I can do binary image without any loss I can't do the skeleton
analysis. Is there some other way to find about the details of cracks ( any
in built plugins in imageJ or Fiji)?

Thanks
Anu

On Sunday, October 15, 2017, anusuya pal <[hidden email]> wrote:

> Hello,
>
> I have an image (gray-scaled of the raw image) and I want to have the
> detailed information about the cracks. Basically, I want to count all its
> junctions, no of node points, branches and measure their average length. I
> have found many threads discussed related to cracks in this forum. What I
> come up is that the appropriate tool for my case will be *Analyze
> Skeleton (2D/3D).*
>
> But prior to use this tool, I need to binarise my image, and here the
> problem comes. As I do that, the cracks do not exactly match with those in
> the original image.
>
> For the convenience, I have attached the image (n_varian.jpg) where I
> applied variance filter just to get the cracks. And then tried to make it
> binary, but it fails.
>
> Are there any suggestions to get the details of the cracks or to solve
> this issue?
>
> Thanks
> Anu
>
>
>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html
Reply | Threaded
Open this post in threaded view
|

Re: Crack details extraction

Swayne, Theresa C.

Dear Anu,

Accurate segmentation (in this case, detecting the cracks) depends on the quality of the original image and also the method used to segment.

For example, to improve the input image quality, it would probably help to correct the background illumination (Process > Subtract Background) before running edge detection.

There are many methods to segment an image, that is, to divide “real” features from background. When you run the “make binary” command, ImageJ applies one *particular* algorithm to threshold the edge-detection image. But there are different algorithms you can try (see http://imagej.net/Auto_Threshold for some). You can also use more sophisticated methods (not only dependent on intensity) – such as trainable Weka segmentation.

There are several useful links on segmentation here:
http://forum.imagej.net/t/segmentation-newbie-help/7053

Hope this helps.
------------------------------------
Theresa Swayne, Ph.D.
Associate Research Scientist
Manager, Confocal and Specialized Microscopy Shared Resource<http://hiccc.columbia.edu/research/sharedresources/confocal>

Herbert Irving Comprehensive Cancer Center
Columbia University Medical Center
1130 St. Nicholas Ave., Room 222A
New York, NY 10032
Phone: 212-851-4613
[hidden email]<mailto:[hidden email]>



From: ImageJ Interest Group <[hidden email]> on behalf of anusuya pal <[hidden email]>
Reply-To: "[hidden email]" <[hidden email]>
Date: Wednesday, October 18, 2017 at 12:41 PM
To: "[hidden email]" <[hidden email]>
Subject: Re: Crack details extraction

Hello,

May I request anyone to look into this matter?
Unless I can do binary image without any loss I can't do the skeleton
analysis. Is there some other way to find about the details of cracks ( any
in built plugins in imageJ or Fiji)?

Thanks
Anu

On Sunday, October 15, 2017, anusuya pal <[hidden email]<mailto:[hidden email]>> wrote:

Hello,

I have an image (gray-scaled of the raw image) and I want to have the
detailed information about the cracks. Basically, I want to count all its
junctions, no of node points, branches and measure their average length. I
have found many threads discussed related to cracks in this forum. What I
come up is that the appropriate tool for my case will be *Analyze
Skeleton (2D/3D).*

But prior to use this tool, I need to binarise my image, and here the
problem comes. As I do that, the cracks do not exactly match with those in
the original image.

For the convenience, I have attached the image (n_varian.jpg) where I
applied variance filter just to get the cracks. And then tried to make it
binary, but it fails.

Are there any suggestions to get the details of the cracks or to solve
this issue?

Thanks
Anu




--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html