ImageJ Straighten: why does marker brightness increase

Hi,

Would anyone please be able to point me in the direction of the theory behind the straightener in ImageJ which is accessible via Edit->Selection->Straighten? I can’t seem to find it myself.

My main question is why the brightness and apparent diffusion of fluorecence increases when straightening an image stack as you can see in an example in the image below.

Thank you so much for any help!

In the image header you see that the resulting image is 32-bit instead of your original 8-bit. This means it has to be displayed differently, thus the brightness contrast settings will be slightly different and the image will just appear to be different in its overall brightness. The change is thus just visual.

To note is that the transformation that you apply to your image will change the intensity values, due to interpolation that has to be done in order to straighten it. Although these changes will be more subtle.

Thank you. Have you come across anywhere that explains the theory of how the straightening occurs at all?

Sorry this refers to a different plugin…
Hm the ImageJ docu is not very extensive about it.
https://imagej.nih.gov/ij/docs/guide/146-27.html#sub:Straighten

But you can dig in the code: https://github.com/imagej/imagej1/blob/master/ij/plugin/Straightener.java

I am sorry that is as far as I can guide you…

It is described in the documentation for the plugin:

https://imagej.nih.gov/ij/plugins/straighten.html

There is a reference to the publication that describes the algorithm. I guess they would also describe the type of interpolation that will be done on the image.

Cheers,
Christopher

1 Like

Hi @est and @schmiedc,

I just happened to read @est 's question and it sparked my curiosity. For what I can gather from the source code (I’m not profficient in Java ), the Straightener plugin first generates a finely spaced spline (1 pixel steps) from the original line roi. Then for each small spline segment, it interpolates using a line perpendicular to the segment (length equal to the roi’s line width, centered in the spline) from the original image to the final result.

Hope this is useful (if it makes any sense at all :grin:)

Cheers,
Nico

2 Likes