feret diameter orienatation

[prosto_w_bemben]

Hello. It's my first post on forum so please understand if You get me wrong. I'm making raindrop splash image analysis in NI Vision Assistant and I would like to get info: is there any option to make max feret diameter measurement in one orientation (for Y axis). In few images I get this parameter but feret diameter oriantation is always in 2-5 degrees. I would like to get this max feret diameter only in horizontal way (0 or 180 degrees), is it possible? Please help me.

[BruceAmmons]

Don't use feret diameter.  Use Max Horiz Segment instead.  Get the Left and Right values and subtract to get length.

Bruce

[prosto_w_bemben]

@bruceammons I was thinking about that, but in NI Vision I only saw "number of horizontal/vertical segments" and average length of horizontal/vertical segment. There isn't max horizontal segment length.

[BruceAmmons]

When I looked at IMAQ Particle Analysis, those properties were available.  They are the last ones in the first section of properties.

If you are using Vision Assistant, those properties may not be available.  This is a case where LabVIEW has more options.

Bruce

[prosto_w_bemben]

Ok thanks, I will check this tomorrow

[prosto_w_bemben]

I checked this out and found this parameter (max horizontal segment length left/right) in both NI Vision and LabView. This could be an idea to measure what I want to, but also I would like to measure other lengths and probably this could be possible only by using max feret diameter.
I attach sample of many suchlike images that I have to analyze.

As You could see I marked two green lines lengths that I would like to know if it is possible to measure them: max crown stretch and max feret diameter in horizontal (this should be upper on picture, but this is only sample). The others lengths (showed by yellow lines) have been already measured in NI Vision. Have You any idea to measure this: max crown stretch and max feret diameter in horizontal?? Maybe both of them could be measured by max feret diameter parameter?

[BruceAmmons]

Running the Convex routine on your particle first might solve some of your problems.  It will fill in the notches so you get a better Max Horiz measurement.

You could use just the X coordinates of the max feret, or use the X coordinates of the particle boundary perhaps.  There are a lot of measurements that you could manipulate to get the data you need.  You just need to know exactly what you are trying to measure.

Bruce

[KevinC]

This discussion might also be interesting to you.

http://forums.ni.com/t5/Machine-Vision/How-Can-I-estimate-the-true-width-of-a-particle/m-p/3090240#M...