Question

在tf.image.central_crop函数的Tensorflow文档中：

Remove the outer parts of an image but retain the central region of
the image along each dimension. If we specify central_fraction = 0.5,
this function returns the region marked with "X" in the below diagram.

 --------
|        |
|  XXXX  |
|  XXXX  |
|        |   where "X" is the central 50% of the image.
 --------

请考虑以下代码：

In [2]: import tensorflow as tf
In [3]: image_raw = tf.placeholder(tf.string)
In [4]: image = tf.image.decode_jpeg(image_raw, channels=3)
In [5]: crop = tf.image.central_crop(image, central_fraction=0.5)
In [6]: init_op = tf.global_variables_initializer()
In [7]: sess = tf.Session()
In [8]: sess.run(init_op)
In [9]: image_np, crop_np = sess.run([image, crop],
   ...:     feed_dict={image_raw: open("/tmp/test.jpg", 'rb').read()})
In [10]: image_np.shape
Out[10]: (456, 450, 3)

原始图片尺寸为456x450

In [11]: crop_np.shape
Out[11]: (228, 226, 3)

裁剪尺寸为228x226

其中面积比为：

In [12]: 228*226 / (456*450.)
Out[12]: 0.2511111111111111

不是我预期的0.5。有人可以帮忙解释一下吗？

Answer 1

正如@Wes Doyle指出的那样，它返回每个维度的50％，这与评论中的插图一致，原始大小为4x8（4＆＃34; |＆＃34;符号和8＆＃34 ; - ＆＃34;标志），裁剪尺寸为2x4（2＆＃34; X＆＃34;垂直，4＆＃34; X＆＃34;水平）。

Tensorflow image.central_crop（mis）行为

1 个答案: