问题描述
我有一些要中心化的图像阵列(减去均值并除以标准差).我可以这样简单吗?
I have some numpy arrays of images that I want to center (subtract the mean and divide by the standard deviation). Can I simply do it like this?
# x is a np array
img_mean = x.mean(axis=0)
img_std = np.std(x)
x = (x - img_mean) / img_std
推荐答案
我不认为这是您要尝试做的事情.
假设我们有一个像这样的数组:
I don't think this is what you are trying to do.
Let's say we have an array like this:
In [2]: x = np.arange(25).reshape((5, 5))
In [3]: x
Out[3]:
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19],
[20, 21, 22, 23, 24]])
x.mean(axis=0)计算每列(轴0)的平均值:
x.mean(axis=0) calculates the mean value for each column (axis 0):
In [4]: x.mean(axis=0)
Out[4]: array([ 10., 11., 12., 13., 14.])
从我们原始的x数组中减去，每个值都减去其列的平均值:
Subtracted from our original x array, each value gets subtracted by its column's mean value:
In [5]: x - x.mean(axis=0)
Out[5]:
array([[-10., -10., -10., -10., -10.],
[ -5., -5., -5., -5., -5.],
[ 0., 0., 0., 0., 0.],
[ 5., 5., 5., 5., 5.],
[ 10., 10., 10., 10., 10.]])
如果未为x.mean指定轴，则整个数组将被获取:
If we don't specify an axis for x.mean, the whole array is being taken:
In [6]: x.mean(axis=None)
Out[6]: 12.0
这就是您一直在使用x.std()的方式，因为对于 np.std 和 np.mean 的默认轴为None.
这可能就是您想要的:
This is what you've been doing with x.std() all the time, since for both np.std and np.mean the default axis is None.
This might be what you want:
In [7]: x - x.mean()
Out[7]:
array([[-12., -11., -10., -9., -8.],
[ -7., -6., -5., -4., -3.],
[ -2., -1., 0., 1., 2.],
[ 3., 4., 5., 6., 7.],
[ 8., 9., 10., 11., 12.]])
In [8]: (x - x.mean()) / x.std()
Out[8]:
array([[-1.6641005, -1.5254255, -1.3867504, -1.2480754, -1.1094003],
[-0.9707253, -0.8320502, -0.6933752, -0.5547002, -0.4160251],
[-0.2773501, -0.1386750, 0. , 0.1386750, 0.2773501],
[ 0.4160251, 0.5547002, 0.6933752, 0.8320502, 0.9707253],
[ 1.1094003, 1.2480754, 1.3867504, 1.5254255, 1.6641005]])
这篇关于将一堆图像居中的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！