How can you mask images that are a different shape than the ones you need?

I am searching an image of an eye for small red blobs, each smaller than 10 pixels in radius. After masking out all non-red pixels, I am left with a sizable number of pixels, some of which are not in the dots I need. Instead, they are in veins, which I do not want to appear in the final image. How do I create a mask for these veins?

[Kirby Fears]

AG,

The problem as stated is hard to answer. Please provide more information to clarify what the data representing a vein (versus a dot) looks like.

Think creatively about the properties of these two different kinds of data. You just need a condition that serves as a catch-all for non-dots.

Let me take a stab at it anyway. You mentioned, "I am searching an image of an eye for small red blobs, each smaller than 10 pixels in radius."

It sounds like a vein (falsely identified as a blob) would have extra red pixels bordering your 10 pixel radius. If that's the case, you just need to exclude blobs that have red pixels bordering the 10 pixel radius.

A rule along these lines should help.