Hello,
I am new to ImageJ and image processing in general, and I'm hoping for some
advice on how best to automate some of part of processing a large
collection of images. The images I'm measuring are cylindrical
cross-sections cut (using a chainsaw) from fallen logs. My goal is to find
the proportion of void space for the set of images. I find the area of the
cylindrical cross section, and then I find the area of any void space w/in
the cross section. Often there is no void space, so in this case I just
measure the area of the cross section. The void space is commonly simply a
hole in the center of the cylinder, however the void space can also take
the form of multiple holes, or might even extend all the way past the edge
of the cylinder (if void space is surrounded by more than 180* of wood it
is void space).
Below are the manual steps that I use to measure an image. These are
simple to execute but can be very tedious. I'm having a hard time seeing
how I can automate the process, since there are two clearly manual steps -
setting the scale, and drawing the cylinder/void space regions. For the
latter, I might be able to look into defining all of my ROIs first, saving
them, and then applying via a batch process. I unsure how I would set an
individual scale for each image though.
Set the scale (will be different for each image)
1. Draw a line the length of one cm (using ruler in photo)
2. Analyze -> Set Scale
a. Known distance = 1
b. Unit: cm
c. Global: check
Find the Area
1. Freehand draw the region to be measured
a. Draw multiple freehand regions by clicking “shift”
2. Edit Selection -> Create Mask
3. File -> Save As -> PNG (to save for later)
a. Use the wand tool and
i. click each black
area
ii. then click
measure
Do this for once the outside cylinder, and once for the void area
I'm happy to supply a few sample images if that would help. I am
comfortable coding in Java if that is required.
Thank you!
Ben
--
ImageJ mailing list:
http://imagej.nih.gov/ij/list.html
Locked
