Dustin!
Thanx again a lot for your help. I used moments based method for my images. It is the only one that seems to work, but i still had to increase the lower threshold by 4-5 gray levels. Anyway, it seems to work.
As to posting it, i think i could do it, even though i am doing it for the company, but am only working as a trainee. right now i have two solution - one hough based, the other just calculates the particle's max intercept and elongation factor, and judging on that it is possible to determine whether or not it is a scratch(line).
Do you know wwhere i can post my sources? i guess here at ni developer zone.
greetings,
andrew