how does zoom out works in data augmentation?
I'm reading a doc on data augmentation in Keras ,and it says that randomZoom(0.2) zooms in and out by a factor in rage of 20%. how does zoom out works? does it add white border to the image?
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
It seems that it mirrors the image in borders; so it is like putting smaller versions of the image around the image(having 9 smaller images together) and then cropping the big image with the size of the original image.