Resampling a Numpy Array Representing an Image

Resampling a numpy array representing an image

Based on your description, you want scipy.ndimage.zoom.

Bilinear interpolation would be order=1, nearest is order=0, and cubic is the default (order=3).

zoom is specifically for regularly-gridded data that you want to resample to a new resolution.

As a quick example:

import numpy as np
import scipy.ndimage

x = np.arange(9).reshape(3,3)

print 'Original array:'
print x

print 'Resampled by a factor of 2 with nearest interpolation:'
print scipy.ndimage.zoom(x, 2, order=0)

print 'Resampled by a factor of 2 with bilinear interpolation:'
print scipy.ndimage.zoom(x, 2, order=1)

print 'Resampled by a factor of 2 with cubic interpolation:'
print scipy.ndimage.zoom(x, 2, order=3)

And the result:

Original array:
[[0 1 2]
 [3 4 5]
 [6 7 8]]
Resampled by a factor of 2 with nearest interpolation:
[[0 0 1 1 2 2]
 [0 0 1 1 2 2]
 [3 3 4 4 5 5]
 [3 3 4 4 5 5]
 [6 6 7 7 8 8]
 [6 6 7 7 8 8]]
Resampled by a factor of 2 with bilinear interpolation:
[[0 0 1 1 2 2]
 [1 2 2 2 3 3]
 [2 3 3 4 4 4]
 [4 4 4 5 5 6]
 [5 5 6 6 6 7]
 [6 6 7 7 8 8]]
Resampled by a factor of 2 with cubic interpolation:
[[0 0 1 1 2 2]
 [1 1 1 2 2 3]
 [2 2 3 3 4 4]
 [4 4 5 5 6 6]
 [5 6 6 7 7 7]
 [6 6 7 7 8 8]]

---

Edit: As Matt S. pointed out, there are a couple of caveats for zooming multi-band images. I'm copying the portion below almost verbatim from one of my earlier answers:

Zooming also works for 3D (and nD) arrays. However, be aware that if you zoom by 2x, for example, you'll zoom along all axes.

data = np.arange(27).reshape(3,3,3)
print 'Original:\n', data
print 'Zoomed by 2x gives an array of shape:', ndimage.zoom(data, 2).shape

This yields:

Original:
[[[ 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]]]
Zoomed by 2x gives an array of shape: (6, 6, 6)

In the case of multi-band images, you usually don't want to interpolate along the "z" axis, creating new bands.

If you have something like a 3-band, RGB image that you'd like to zoom, you can do this by specifying a sequence of tuples as the zoom factor:

print 'Zoomed by 2x along the last two axes:'
print ndimage.zoom(data, (1, 2, 2))

This yields:

Zoomed by 2x along the last two axes:
[[[ 0  0  1  1  2  2]
  [ 1  1  1  2  2  3]
  [ 2  2  3  3  4  4]
  [ 4  4  5  5  6  6]
  [ 5  6  6  7  7  7]
  [ 6  6  7  7  8  8]]

 [[ 9  9 10 10 11 11]
  [10 10 10 11 11 12]
  [11 11 12 12 13 13]
  [13 13 14 14 15 15]
  [14 15 15 16 16 16]
  [15 15 16 16 17 17]]

 [[18 18 19 19 20 20]
  [19 19 19 20 20 21]
  [20 20 21 21 22 22]
  [22 22 23 23 24 24]
  [23 24 24 25 25 25]
  [24 24 25 25 26 26]]]

Numpy Resize/Rescale Image

Yeah, you can install opencv (this is a library used for image processing, and computer vision), and use the cv2.resize function. And for instance use:

import cv2
import numpy as np

img = cv2.imread('your_image.jpg')
res = cv2.resize(img, dsize=(54, 140), interpolation=cv2.INTER_CUBIC)

Here img is thus a numpy array containing the original image, whereas res is a numpy array containing the resized image. An important aspect is the interpolation parameter: there are several ways how to resize an image. Especially since you scale down the image, and the size of the original image is not a multiple of the size of the resized image. Possible interpolation schemas are:

• INTER_NEAREST - a nearest-neighbor interpolation
• INTER_LINEAR - a bilinear interpolation (used by default)
• INTER_AREA - resampling using pixel area relation. It may be a preferred method for image decimation, as it gives moire’-free
  results. But when the image is zoomed, it is similar to the
  INTER_NEAREST method.
• INTER_CUBIC - a bicubic interpolation over 4x4 pixel neighborhood
• INTER_LANCZOS4 - a Lanczos interpolation over 8x8 pixel neighborhood

Like with most options, there is no "best" option in the sense that for every resize schema, there are scenarios where one strategy can be preferred over another.

Resample and resize numpy array

Instead of passing a single number to the zoom parameter, give a sequence:

scipy.ndimage.zoom(x, zoom=(1.5, 2.), order=1)
#array([[0, 0, 1, 1, 2, 2],
#       [2, 2, 3, 3, 4, 4],
#       [4, 4, 5, 5, 6, 6],
#       [6, 6, 7, 7, 8, 8]])

With the sequences (2., 2.75) and (2., 3.5) you will get output arrays with shapes (6, 8) and (6, 10), respectively.

numpy: Resize 3d array with interpolation in 2d

I think you can do what you want with scipy.ndimage.map_coordinates:

import numpy as np
import scipy.ndimage.interpolation

def resize_batch(image_batch, new_width, new_height):
    image_batch = np.asarray(image_batch)
    shape = list(image_batch.shape)
    shape[1] = new_width
    shape[2] = new_height
    ind = np.indices(shape, dtype=float)
    ind[1] *= (image_batch.shape[1] - 1) / float(new_width - 1)
    ind[2] *= (image_batch.shape[2] - 1) / float(new_height - 1)
    return scipy.ndimage.interpolation.map_coordinates(image_batch, ind, order=1)

print(resize_batch(np.zeros([10, 20, 30]), 60, 15).shape)
# (10, 60, 15)

print(resize_batch(np.zeros([10, 20, 30, 3]), 60, 15).shape)
# (10, 60, 15, 3)

EDIT:

Here are a couple of other versions. This one uses only NumPy operations without SciPy, computing bilinear interpolation "by hand":

import numpy as np

def resize_batch_np(image_batch, new_width, new_height):
    dtype = image_batch.dtype
    n, width, height = image_batch.shape[:3]
    extra_dims = image_batch.ndim - 3
    w = np.linspace(0, width - 1, new_width, dtype=dtype)[:, np.newaxis]
    h = np.linspace(0, height - 1, new_height, dtype=dtype)
    nn = np.arange(n, dtype=np.int32)[:, np.newaxis, np.newaxis]
    ii_1 = w.astype(np.int32)
    ii_2 = (ii_1 + 1).clip(max=width - 1)
    w_alpha = w - ii_1
    w_alpha = w_alpha.reshape(w_alpha.shape + (1,) * extra_dims)
    w_alpha_1 = 1 - (w_alpha)
    jj_1 = h.astype(np.int32)
    jj_2 = (jj_1 + 1).clip(max=height - 1)
    h_alpha = h - jj_1
    h_alpha = h_alpha.reshape(h_alpha.shape + (1,) * extra_dims)
    h_alpha_1 = 1 - (h_alpha)
    out_11 = image_batch[nn, ii_1, jj_1]
    out_12 = image_batch[nn, ii_1, jj_2]
    out_21 = image_batch[nn, ii_2, jj_1]
    out_22 = image_batch[nn, ii_2, jj_2]
    return ((out_11 * h_alpha_1 + out_12 * h_alpha) * w_alpha_1 +
            (out_21 * h_alpha_1 + out_22 * h_alpha) * w_alpha)

And this other one does the same but with Numba:

import numpy as np
import numba as nb

@nb.njit(parallel=True)
def resize_batch_nb(image_batch, new_width, new_height):
    dtype = image_batch.dtype
    n, width, height = image_batch.shape[:3]
    extra_dims = image_batch.ndim - 3
    w = np.empty(new_width, dtype=dtype)
    for i in range(new_width):
        w[i] = (width - 1) * i / (new_width - 1)
    h = np.empty(new_height, dtype=dtype)
    for i in range(new_height):
        h[i] = (height - 1) * i / (new_height - 1)
    ii_1 = w.astype(np.int32)
    ii_2 = np.minimum(ii_1 + 1, width - 1)
    w_alpha = w - ii_1
    w_alpha_1 = 1 - (w_alpha)
    jj_1 = h.astype(np.int32)
    jj_2 = np.minimum(jj_1 + 1, height - 1)
    h_alpha = h - jj_1
    h_alpha_1 = 1 - (h_alpha)
    out = np.empty((n, new_width, new_height) + image_batch.shape[3:], dtype=dtype)
    for idx in nb.prange(n):
        for i in nb.prange(new_width):
            for j in nb.prange(new_height):
                out_11 = image_batch[idx, ii_1[i], jj_1[j]]
                out_12 = image_batch[idx, ii_1[i], jj_2[j]]
                out_21 = image_batch[idx, ii_2[i], jj_1[j]]
                out_22 = image_batch[idx, ii_2[i], jj_2[j]]
                out_1 = out_11 * h_alpha_1[j] + out_12 * h_alpha[j]
                out_2 = out_21 * h_alpha_1[j] + out_22 * h_alpha[j]
                out[idx, i, j] = out_1 * w_alpha_1[i] + out_2 * w_alpha[i]
    return out

The result is the same as before:

import numpy as np

np.random.seed(100)
image_batch = np.random.rand(100, 200, 300, 3).astype(float)
new_width = 60
new_height = 80
out = resize_batch(image_batch, new_width, new_height)
out_np = resize_batch_np(image_batch, new_width, new_height)
out_nb = resize_batch_nb(image_batch, new_width, new_height)
print(np.allclose(out, out_np))
# True
print(np.allclose(out, out_nb))
# True

But the performance improves significantly:

%timeit resize_batch(image_batch, new_width, new_height)
# 211 ms ± 9.36 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
%timeit resize_batch_np(image_batch, new_width, new_height)
# 106 ms ± 1.8 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)
%timeit resize_batch_nb(image_batch, new_width, new_height)
# 48.3 ms ± 142 µs per loop (mean ± std. dev. of 7 runs, 10 loops each)

Is there a better way to resize an image that is in the form of a numpy array?

To avoid having to convert between tf.tensor and np.ndarray tensor data types, you can use skimage (scikit-image) to resize the image directly on np.ndarray objects using the following code segment:

import skimage.transform

kwargs = dict(output_shape=self._size, mode='edge', order=1, preserve_range=True)
im = skimage.transform.resize(im, **kwargs).astype(im.dtype)

To install skimage, follow the installation instructions here: https://pypi.org/project/scikit-image/. Hope this helps!

Use cv2.imshow() to display an np.array

opencv has BGR channel ordering, PIL and matplotlib use RGB order

try not to mix different libraries with different paradigms

Numpy Upsample np.array / increase size of np.array by adding mean value of cosecuitive elements

np.interp:

a = np.array([1,2,3,4,5])
ur = 2  # upsample rate
np.interp(np.arange((len(a)-1)*ur+1)/ur, xp=np.arange(len(a)), fp=a)
# output: array([1. , 1.5, 2. , 2.5, 3. , 3.5, 4. , 4.5, 5.])

---

Related Topics