Hi !
thank you for the ideas,
the blur\sharpen thing is a really nice trick but i think not enough thanks cod65.
wonder what can be achieve with the convolve…
blob tracking can report the size of a ‘blob’. Might be too computationally costly, but there might be something there. I opened it up and messed with it a bit, but forgot how it works for the most part But I do remember it can report the dimensions of contiguous areas.
But what convolution kernel would you use? I think I don’t understand the convolution idea if it’s not a blur you would want to achieve via convolution.(which indeed would be much more slow)
well yes I know what convolution is, it’s just that i think you plan to use the convolution top with a kernel that creates something like a running average filter, essentially creating a blur. If that’s not the idea, then I don’t get it