Hi
I have been looking at Anthony's example of Test or Caption removal and I don't under stand it too well.
Can anyone here explain it to me a little better?
I see in the first step I need a mask. But I don't understand how that is made. Do I have to type in what the text says in order for it to be Edge detected or what?
It looks to me like he wrote the text on the picture before he started and it would be easier to remove this way. I would like it explained using say a picture off the net with a caption on it with an unknown color and size. I realize if the text is too big in size the blur system may not work very well to fill in the missing text but it still sounds like an interesting procedure.