Reputation: 3854
I was used skimage package to rotate an image that has mode edge option
image_rotate = skimage.transform.rotate(image, angle,mode='edge')
mode : {‘constant’, ‘edge’, ‘symmetric’, ‘reflect’, ‘wrap’}, optional Points outside the boundaries of the input are filled according to the given mode. Modes match the behaviour of numpy.pad.
Right now, I am using tensorflow function
tf.contrib.image.rotate(
images,
angles,
interpolation='NEAREST'
)
How can I use tf.contrib.image.rotate
to obtain the result as mode='edge'
in the skimage package? Thanks
Upvotes: 0
Views: 445
Reputation: 937
tf.contrib.image.rotate
does not implement padding. If you'd like to pad your image before rotating it, you'll have to do that yourself using tensorflow operations.
Upvotes: 2