skimage2.measure.block_reduce

skimage2.measure.block_reduce(image, block_size=2, func=<function sum>, cval=0, func_kwargs=None)

Downsample image by applying function func to local blocks.

This function is useful for max and mean pooling, for example.

Parameters:

image(M[, …]) ndarray

N-dimensional input image.

block_sizearray_like or int

Array containing down-sampling integer factor along each axis. Default block_size is 2.

funccallable

Function object which is used to calculate the return value for each local block. This function must implement an axis parameter. Primary functions are numpy.sum, numpy.min, numpy.max, numpy.mean and numpy.median. See also func_kwargs.

cvalfloat

Constant padding value if image is not perfectly divisible by the block size.

func_kwargsdict

Keyword arguments passed to func. Notably useful for passing dtype argument to np.mean. Takes dictionary of inputs, e.g.: func_kwargs={'dtype': np.float16}).

Returns:

imagendarray

Down-sampled image with same number of dimensions as input image.

Examples

>>> from _skimage2.measure import block_reduce
>>> image = np.arange(3*3*4).reshape(3, 3, 4)
>>> image
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, 25, 26, 27],
        [28, 29, 30, 31],
        [32, 33, 34, 35]]])
>>> block_reduce(image, block_size=(3, 3, 1), func=np.mean)
array([[[16., 17., 18., 19.]]])
>>> image_max1 = block_reduce(image, block_size=(1, 3, 4), func=np.max)
>>> image_max1
array([[[11]],
       [[23]],
       [[35]]])
>>> image_max2 = block_reduce(image, block_size=(3, 1, 4), func=np.max)
>>> image_max2
array([[[27],
        [31],
        [35]]])