很多時候我們需要對影象進行區域性對比度歸一化，比如分塊CNN的預處理階段。theano對此提供了一些比較方便的操作。

區域性歸一化的一種簡單形式為：

其中μ和σ分別為區域性（例如3x3的小塊）的均值和標準差。

利用程式碼說明一下如何實現：

import theano
import numpy
from theano.sandbox import neighbours
from theano import tensor as T
from theano import function  
from skimage import io
import time

patch = io.imread('
 
test.jpg', True)
patch_batch = []
batch_size = 384

# 這裡模擬批量處理多個影象塊
for i in xrange(batch_size):
    patch_batch.append(patch)

_input = T.tensor3('_input')

norm_input = T.reshape(_input, T.concatenate([(batch_size,1),_input.shape[1:]]), ndim=4)

# neighbours.images2neibs(ten4, neib_shape, neib_step=None, mode='valid')，其中neib_shap為鄰域大小，這裡設定為3x3，
 

# neib_step指定步長，我們需要對每一個點進行歸一化，因此設定為(1x1)
# 該函式返回一個shape為(batch_size x patch_width x patch_height, 9)的陣列，每一行代表每個點的9個鄰居
neibs = neighbours.images2neibs(norm_input, (3, 3), (1, 1), 'wrap_centered')
 
_means = T.mean(neibs, axis=1)
_stds = T.std(neibs, axis=1)

_flatten_input = _input.flatten()
normed_result 
 
= (_flatten_input - _means) / (_stds + 1 / 255)

# 將結果重新reshape為影象大小
reshpaed_normed_result = T.reshape(normed_result, _input.shape)
                             
shared_patch = theano.shared(numpy.asarray(patch_batch, dtype=theano.config.floatX), borrow=True)
calc_norm = function([], reshpaed_normed_result, givens={_input:shared_patch})

start_time = time.clock()
dd = calc_norm()
end_time = time.clock()

print end_time - start_time