Particle Analysis - Do they make a line?

Hi all,

We are doing a far amount of image analysis mostly on a micro level. We have identified features/defects and measured size and position.

The next quest is did the defects arise from a macro level event? So the first test I would like to do is do the defects form a line. The simplistic regression analysis I can handle, but what I am thinking about is how to set the range of defects to consider. I have a large wafer which multiple defects that show up as dots on a macro level and the eye can logically group them into separate line defects.

How do I think about programmically grouping them into separate lines and then creating some figures of merit?

I hope I am clear in my request.
The typical way to handle this is using the Hough transform.

It's even implemented in Igor. See DisplayHelpTopic "Hough Transform"
Thank you very much.

This is the place to start and I knew Igor would have the tools built in.