Image interpolation in SciPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of resizing an image with scipy.ndimage.zoom

What is the shape of the output image after applying scipy.ndimage.zoom with a zoom factor of 2 on a 2D grayscale image of shape (50, 50)?

SciPy

import numpy as np
from scipy.ndimage import zoom

image = np.ones((50, 50))
zoomed_image = zoom(image, 2)
print(zoomed_image.shape)

A (51, 51)

B (50, 50)

C (25, 25)

D (100, 100)

💡 Hint

Think about how zoom factor affects each dimension size.

The zoom factor of 2 doubles each dimension size, so (50, 50) becomes (100, 100).

🧠 Conceptual

intermediate

Understanding interpolation order in scipy.ndimage.zoom

Which interpolation order in scipy.ndimage.zoom corresponds to cubic spline interpolation?

A order=1

B order=3

C order=0

D order=5

💡 Hint

Order 0 is nearest neighbor, order 1 is linear.

Order 3 corresponds to cubic spline interpolation in scipy.ndimage.zoom.

❓ data_output

advanced

Resulting pixel value after zoom with linear interpolation

Given a 1D array arr = np.array([0, 10, 20]), what is the output of scipy.ndimage.zoom(arr, 2, order=1)?

SciPy

import numpy as np
from scipy.ndimage import zoom

arr = np.array([0, 10, 20])
result = zoom(arr, 2, order=1)
print(result)

A [ 0. 5. 10. 15. 20.]

B [ 0. 10. 20.]

C [ 0. 7.5 15. 20.]

D [ 0. 3.3 6.6 10. 20.]

💡 Hint

Linear interpolation fills values evenly between points.

Zoom with factor 2 and linear interpolation doubles points, interpolating linearly between original values.

🔧 Debug

advanced

Identify the error in zooming a color image

What error occurs when running this code to zoom a color image with shape (100, 100, 3) using scipy.ndimage.zoom(image, 2)?

SciPy

import numpy as np
from scipy.ndimage import zoom

image = np.ones((100, 100, 3))
zoomed = zoom(image, 2)
print(zoomed.shape)

A ValueError: zoom() missing required argument 'zoom'

B TypeError: float object cannot be interpreted as an integer

C Output shape is (200, 200, 6) which is incorrect

D Output shape is (200, 200, 3) which is correct

💡 Hint

Check how zoom interprets a single float for multi-dimensional arrays.

Passing a single float applies zoom to all axes, including color channels, doubling channels from 3 to 6.

🚀 Application

expert

Choosing correct zoom parameters for RGB image resizing

You want to resize an RGB image of shape (120, 160, 3) to double its width and height but keep color channels unchanged using scipy.ndimage.zoom. Which zoom parameter is correct?

A (2, 2, 1)

B 2

C (2, 2, 2)

D (1, 1, 3)

💡 Hint

Color channels should not be scaled.

Zoom factor must be a tuple matching array dimensions; spatial axes doubled, channel axis kept 1.