Batch quantify black levels / output histogram data

[ArchAnGeL]

Ok, bit of a strange request this one but here goes.

I have a series of B&W images (several thousand) and need to quantify the amount of black in each one

They are basically CT scans so I am thinking I need the histogram areas split into 0 to 255 or say the total area between say 0 and 155.

This needs to be per image and output into a common file e.g csv, xls etc so I can quantify it.

I know this is possible manually so hopefully someone has an idea on how to automate/batch process this

Thanks

 

[ArchAnGeL]

If it help or jogs any memories, I can see how this would be possible in imageJ but I don't understand the macro language

 

[D.P.]

A quick openCV program wouldn't be hard if you can program. Otherwise I know the gimp would allow you to make a script that would do this.

If I wasn't so busy would throw something together myself.

 

[ArchAnGeL]

Thanks for the pointers. Probably more confused by gimp than imagej at this stage

 

[D.P.]

from PIL import Image
import glob, os


for f in glob.glob("*.jpg"):
   im = Image.open(f)
   hist = im.histogram()
   count = hist[0]    # Assume BW image, hist is a list of 256 values, entry 0 is the count for black pixels.
   print "Image ", f, " has [", count, "] many pixels that are all black"

Untested, but that is all you would need in theory in python. Most usable operating systems (linux, unix, MacOSX, BSD) will have python installed at the system level but you might have to install it on windows unless Microsoft have caught up. You will have to install the PIL library, on anything unixy:

pip install PIL

 

[ArchAnGeL]

So is that running just in python or using python to script another prog?

 

[D.P.]

Pure python, just using the PIL library (which I assume is a native c lib)

 

[ArchAnGeL]

thankyou