Python PIL | Image.resize() method

The Image.resize() method in Python's PIL (Pillow) library is used to change the size of an image. It creates a new resized copy without modifying the original image. This method is mainly used when preparing images for thumbnails, preprocessing images for ML models, or standardizing image dimensions.

Note: For this article we will use a sample image "bear.png", to download click here.

Example: Here's a example to resize an image to 300×300, and displays the resized output.

from PIL import Image

img = Image.open("bear.png")
res = img.resize((300, 300))
res.show()

Output

Explanation:

• Image.open() loads the original image
• resize((300, 300)) creates a resized copy with the given width and height
• show() displays the resized image

Syntax

Image.resize(size, resample=Image.BICUBIC)

Parameters:

• size: A tuple (width, height) defining the new image size.
• resample (Optional): filter for quality. Common options are: Image.NEAREST -> Fast, lowest quality, Image.BILINEAR -> Moderate quality, Image.BICUBIC -> High quality (default) and Image.LANCZOS -> Best quality for downscaling

Examples

Example 1: This example resizes the image to a fixed square size suitable for thumbnails.

from PIL import Image

img = Image.open("bear.png")
out = img.resize((200, 200), Image.LANCZOS)
out.show()

Output

Explanation: resize((200, 200), Image.LANCZOS) applies high-quality downscaling for a sharper result.

Example 2: Here, we change only the width and automatically compute the height to maintain aspect ratio.

from PIL import Image

img = Image.open("bear.png")
w, h = img.size
new_w = 400
new_h = int(h * (new_w / w))
out = img.resize((new_w, new_h), Image.BICUBIC)
out.show()

Output

Explanation:

• new_h = int(h * (new_w / w)) calculates proportional height
• resize((new_w, new_h)) resizes while keeping aspect ratio

Example 3: This example uses NEAREST to create a pixelated effect, often used for retro-style graphics.

from PIL import Image

img = Image.open("bear.png")
out = img.resize((120, 120), Image.NEAREST)
out.show()

Output

Explanation: Image.NEAREST enlarges or shrinks the image using the nearest pixel, creating a blocky effect.