Detecting Striations


#1

I am trying to see if it is possible for ImageJ to reliably detect striations. So far, I have tried the ridge detection plugin and the measure directionality feature without much success. The ridge detection plugin seemed to put too much emphasis on joining lines together while measure directionality seemed to catch lines that did not exist. I was hoping someone else has some suggestions I could try out. Thanks in advanced.


#2

Good day,

would you please be so kind and post a typical image because otherwise it appears near to impossible to help.

Furthermore, it would help to know what you mean by “reliably detect striations”, i.e. what exactly is the goal (output) of your analyses (enhanced images of what kind, numerical data of what kind)?

Regards

Herbie


#3

DSCN8546-1

This is a typical image that I have been analyzing. It was taken using a microscope. There are horizontal striations running throughout the material. I am hoping ImageJ could detect and highlight the striations in an image with some consistency. If ImageJ can produce any numerical data, that would likely be an added bonus.

My experiments with the Ridge Detection plugin provided the following result
DSCN8546-1 Width 5 Upper Threshold 8
I am hoping that ImageJ could correctly determine and highlight the striations in a similar manner to how ImageJ has detected the ridges in this image. The problem with this image is that it appears the plugin is focused on creating connected lines rather than detecting the straight striations.

Thanks.


#4

Good day,

and thanks for the images and details. The task appears clear now but I must admit that it won’t be easy to come up with satisfying results.

Here is an analysis of the orientation distribution of your sample image:

The plot starts with the horizontal orientation and the orientation angle proceeds counter-clockwise.

It is evident, that the vertical structures in your image are at least as dominant as the more or less horizontal ones that you are interested in. Any analysis will thus be disturbed by the vertical structures. Consequently, you may try to suppress the vertically oriented structures first.

Finally, it would be highly desirable to see images with considerably higher spatial resolution.

HTH a bit

Herbie


#5

Thanks for your help so far. I will have to discuss the viability of suppressing certain structures before doing the image analysis. In the mean time, here is the original image that the analyzed sample was taken from.


#6

Hello.
In the sampled image the image section looks fairly evenly illuminated therefore have you tried segmentation?
Its east to try and might give you the results you are looking for.
Best on Luck

Bob