Particle size measurement

Image Analysis
#1

Dear All,

I have particle sizes and area fractions that I need to measure in my metallographic samples. An example is attached below. Unfortunately the colours of the particles and the matrix do not show great contrast. Due to my very imited skills in Image J, I have a hard time in segmenting the particles so that Image J can succesfully discriminate and measure them. The best advice I got is to print the pictures, manually outline the particles with distinguishable colours, and rescan them; however I don’t think it is a good procedure.

I was wondering if someone could help me with an image processing procedure I could follow (plug-ins to use, etc) to obtain a sufficient segmentation for succesful measurements.

Thank you very much.

inlenstest01.tif (867.5 KB)

#2

You could try applying some filters and/or other morphological operations on the data. When I tried to threshold your image, I noticed there seemed to be a gradient in the illumination of the sample. I used Process > Subtract Background... with a ball radius of 100, dark background, smoothing disables, and sliding parabola enabled. This seemed to improve but not entirely solve the thresholding issue. It definitely seemed to improve the thresholding of the edges.