How to apply a voronoi on regions rather than centroids

Hi there,

I have some miroscopy images of particles that are already segmented to a mask . Now i want to go iteratively through all particles to find all its neighbours and their closest distance (see attached image 2).

I know that i could identify all particle centroids with regionprobs and then do a voronoi construction on the centroids to find all neighboring centroids and the distance from centroid to centroid, but I'm after the distance between the perimeters not the centers.

Is there any function that I could use? Basically doing a voronoi starting at the perimeters of the particles NOT their centers to identify which particles are adjasent to each other? If i have this information, I could use pdist to find the minium distance between the perimeters of a pair of particles..

thanks a lot for help

Best wishes

Answer 1

[Image Analyst]

It sounds like you're wanting the skiz. Search the help for skiz and you'll find a demo. Also look at these

Watershed transform question from tech support » Watershed transform question from tech support

watershed

https://www.mathworks.com/help/images/marker-controlled-watershed-segmentation.html

[Hannes Frey]

This is exaclty what I'm after. Thanks so much! This is great