region of interest

[Tentotheminus9]

Hi,

I've got a VI that takes an image, applies a threshold, and then applies a centroid. This gives me the x and y position of a particle in the image. However, I've recently started to use more 'messy' images with lots of rubbish around the particle. I've been trying to apply a ROI to the image. I've put it after the threshold, and then sent the data to the centroid. This doesn't seem to work, as the threshold will only work when i've got the image set to 16-bit, but the image mask for the ROI needs to be 8-bit. I've tried converting etc but I can't quite get it to work. When I set it so that the threshold works, then the x,y positions cease to work form the centroid.

Basically I want to load an image, specify (manually is fine for the time being) the coordinates of the ROI, and then apply the centroid just to that ROI, thus cutting out all of the rubbish around the particle.

Please find attached the VI.

Thanks,

Rob.

[standrews]

Hey man,

Judging by your other post it seems like you are trying to do a LabVIEW version of the Crocker/Weeks/Dufresne tracker routine.

We've already done it using pattern matching instead of thresholding. www.st-and.ac.uk/~gfm2/tracker.htm

I dont think you can get around the 8-bit limitations of the inherent vision VIs.

The tracker you can download from out site is limited because of this too. I think you have to develop your own VIs to substitute the NI ones if you want to use more complex images.

I have also written a 'blob analysis' tracker that will be up on the site soon. Check it out, could save you a lot of time.

Cheers,

Graham