You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Would it be possible to do masked MinMaxDifference where black area from a mask is omitted?
That would allow to get stats from only the relevant parts of the frames.
The text was updated successfully, but these errors were encountered:
I suppose it should work in general for all min/max/median/... functions. Thresholded modes (median as well, which is an 50% threshold) however calculate with with total pixel count.
Question: if - special case - mask is all black, then the total pixel count is zero. This would affect average as well.
What is the required action for this? Report an error?
Question2: the full opacity (255 in 8-bits) can also have special meaning. Does your proposal have a reason to use 0-black for the purpose? (I don't mind, just asking, I don't know the actual use case)
Would it be possible to do masked MinMaxDifference where black area from a mask is omitted?
That would allow to get stats from only the relevant parts of the frames.
The text was updated successfully, but these errors were encountered: