Image interpolation in SciPy - Mini Project: Build & Apply

Image interpolation

📖 Scenario: You have a small grayscale image represented as a 2D array of pixel values. You want to make the image bigger smoothly by estimating new pixel values between the existing ones. This process is called image interpolation.Imagine you took a small photo and want to print it larger without it looking blocky or pixelated.

🎯 Goal: Use scipy to perform image interpolation and enlarge the image smoothly.

📋 What You'll Learn

Create a 2D numpy array called small_image with exact pixel values

Create a variable called zoom_factor to control the enlargement

Use scipy.ndimage.zoom with order=3 to interpolate the image

Print the shape of the original and the enlarged image

💡 Why This Matters

🌍 Real World

Image interpolation is used in photo editing, medical imaging, and satellite image processing to enlarge images without losing quality.

💼 Career

Understanding image interpolation is important for data scientists working with computer vision, image analysis, and machine learning tasks involving images.

1

Create the small image array

Create a 2D numpy array called small_image with these exact pixel values: [[10, 20, 30], [40, 50, 60], [70, 80, 90]].

SciPy

import numpy as np
# Create the 2D numpy array small_image with the given pixel values
# Your code here

Need a hint?

Use np.array to create a 2D array with the exact values.

2

Set the zoom factor

Create a variable called zoom_factor and set it to 2 to enlarge the image by 2 times.

SciPy

import numpy as np
small_image = np.array([[10, 20, 30], [40, 50, 60], [70, 80, 90]])
# Set zoom_factor to 2
# Your code here

Need a hint?

Just assign the number 2 to the variable zoom_factor.

3

Interpolate the image using scipy

Import scipy.ndimage and use scipy.ndimage.zoom with order=3 to create a new variable large_image by zooming small_image by zoom_factor.

SciPy

import numpy as np
small_image = np.array([[10, 20, 30], [40, 50, 60], [70, 80, 90]])
zoom_factor = 2
# Import scipy.ndimage and create large_image by zooming small_image with order=3
# Your code here

Need a hint?

Use scipy.ndimage.zoom with order=3 for smooth cubic interpolation.

4

Print the shapes of original and enlarged images

Print the shape of small_image and the shape of large_image using two separate print statements.

SciPy

import numpy as np
import scipy.ndimage
small_image = np.array([[10, 20, 30], [40, 50, 60], [70, 80, 90]])
zoom_factor = 2
large_image = scipy.ndimage.zoom(small_image, zoom=zoom_factor, order=3)
# Print the shapes of small_image and large_image
# Your code here

Need a hint?

Use print(small_image.shape) and print(large_image.shape) to show the sizes.