recently I started to work on automatically analyzing the type of picture which is attached to this post 202010210005.tif. Unfortunately “automatically” is far from what I achieved so far
I’m not sure if my questions are too basic or the project is out of any scope but anyway I appreciate any kind of help. I think an automatization is then possible when I’ve solved all problems so far.
The image you see is a structure made out of glass in the region of a few µm and the dimensions are needed to characterize it.
In the picture you see six capillaries inside a hexagon. I want to extract various dimensions from it. Ideally by using a scale bar to connect it to a dimension like µm. I would always prefer ellipses to circles because they contain more information.
-> the ellipse/circle the six capillaries enclose; I want it from the outer side of the wall not from the middle even if it just a few pixels thick; or at least I want to be able to determine on which side of the wall the capillary is measured
-> inner and outer size of the elliptical capillaries (= inner/outer long and short axis); to get a precise value for the wall thickness another picture is used which shows the wall itself in a higher magnification so this is more important due to the previous step
-> the ellipse/circle the six corners of the hexagon are forming
The importance is in descending order.
What have you tried already?
So far I was able to extract a “mean” value of the six capillaries by using a watershed method combined with an “Ellipse Split”. That means the ellipse is drawn through the middle of the thin line which represents the wall of the capillary. (1.1 MB) 202010210005-watershed-lines-ellipse-fit.tif (1.1 MB)
What stops you from proceeding?
Unfortunately I have no idea how to get the inner and outer diameter of these capillaries, the enclosed diameter and the hexagon.
Hopefully somebody is willed to help me and that there is a solution to my task.
Thanks and greetings,