Function Differences with tf.image.rot90

View Source On Gitee

tf.image.rot90

tf.image.rot90(
    image,
    k=1,
    name=None
)

For more information, see tf.image.rot90.

mindspore.dataset.vision.Rotate

class mindspore.dataset.vision.Rotate(
    degrees,
    resample=Inter.NEAREST,
    expand=False,
    center=None,
    fill_value=0
)

For more information, see mindspore.dataset.vision.Rotate.

Differences

TensorFlow: Rotate the image counter-clockwise, by 90 degrees each time.

MindSpore: Rotate the image counter-clockwise with any specified angle, and pad the area outside the rotated image.

Code Example

# The following implements Rotate with MindSpore.
import numpy as np
import mindspore.dataset as ds

image = np.array([[[0.1], [0.2]], [[0.3], [0.4]], [[0.5], [0.6]]])
result = ds.vision.Rotate(90)(image)
print(result)
# [[[0. ], [0. ]],
#  [[0.4], [0.6]],
#  [[0.3], [0.5]]]

# The following implements rot90 with TensorFlow.
import tensorflow as tf

image = tf.constant([[[0.1], [0.2]], [[0.3], [0.4]], [[0.5], [0.6]]])
result = tf.image.rot90(image, k=1)
print(result)
# [[[0.2], [0.4], [0.6]],
#  [[0.1], [0.3], [0.5]]]