IMAQ setoffset not updating actual position of masked areas

[comrade]

I noticed that IMAQ SetOffset works well for the masking operation (e.g. translationally aligning a mask to an image where the areas to be masked may be some pixels out of place due to incomplete positioning during image aquisition), however the position of the areas that are not masked (pixels >0) are not moved at all, as can be observed in the attached example.

The bounding rectangles of the non-masked areas have to be transformed in addition to performing IMAQ SetOffset otherwise using the particle information of the image mask might not yield the expected results as they d not seem to be altered according to the offset.

Is this behaviour by design or is it unintended?

[jSturm]

Hey comrade,

you still have problems now with the IMAQ SetOffset.vi?

The offset you create with the IMAQ SetOffset VI defines the position of an image mask in RELATION to an other image.

I think you can look at some LabView examples, who explains the field of image masks quite good.
So look at: help/examples: Image Mask.vi & Particle Analyse.vi

Maybe that will help you.

Jerry

[comrade]

Hi jSturm,

to clarify the reason for my initial post here:

It's not so much not being able to cope with the function. I just noticed the fact, that the mask doesn't seem to be actually offsetted but that there is rather additional information stored with the image that says, when using the image, to offset the pixel information according to the given offset (incidently with respect to a coordinate system, not in relation to another image - as there is no other image information that can be passed into the IMAQ SetOffset function). That might be a reasonable explanation for the offset working perfectly as expected while using the image as a mask but the particle information of the offsetted image only yielding the original position values, as if the image had never been offsetted.

To me, this seems a bit of a twofold behaviour which is not directly comprehensible. Therefore I wanted to ask whether there is a particular reason for having the function behave as it does. Of course there may be well-reasoned wisdom behind it. So I thought someone could enlighten me.

[jSturm]

Hey commrade,

so I´m sorry just sending you this examples. 😄 Now I understand your post accurate.
I also test this behaviour and your right the offset working perfectly as expected while using the image as a mask but you can´t see the offset of the pixels. (So where is the offset finally stored?)

So this is truly an interesting point. Of course I have also no real answer for this, but I´m not sure if someone will answer this, because you know the LV code is not open to public.

So just we just wait. 😉