How do I resize an image using PIL and maintain its aspect ratio?

If you are trying to maintain the same aspect ratio, then wouldn’t you resize by some percentage of the original size?

For example, half the original size

half = 0.5
out = im.resize( [int(half * s) for s in im.size] )

This script will resize an image (somepic.jpg) using PIL (Python Imaging Library) to a width of 300 pixels and a height proportional to the new width. It does this by determining what percentage 300 pixels is of the original width (img.size[0]) and then multiplying the original height (img.size[1]) by that percentage. Change "basewidth" to any other number to change the default width of your images.

from PIL import Image

basewidth = 300
img = Image.open('somepic.jpg')
wpercent = (basewidth/float(img.size[0]))
hsize = int((float(img.size[1])*float(wpercent)))
img = img.resize((basewidth,hsize), Image.Resampling.LANCZOS)
img.save('somepic.jpg')

I also recommend using PIL’s thumbnail method, because it removes all the ratio hassles from you.

One important hint, though: Replace

im.thumbnail(size)

with

im.thumbnail(size,Image.ANTIALIAS)

by default, PIL uses the Image.NEAREST filter for resizing which results in good performance, but poor quality.

My ugly example.

Function get file like: “pic[0-9a-z].[extension]”, resize them to 120×120, moves section to center and save to “ico[0-9a-z].[extension]”, works with portrait and landscape:

def imageResize(filepath):
    from PIL import Image
    file_dir=os.path.split(filepath)
    img = Image.open(filepath)

    if img.size[0] > img.size[1]:
        aspect = img.size[1]/120
        new_size = (img.size[0]/aspect, 120)
    else:
        aspect = img.size[0]/120
        new_size = (120, img.size[1]/aspect)
    img.resize(new_size).save(file_dir[0]+'/ico'+file_dir[1][3:])
    img = Image.open(file_dir[0]+'/ico'+file_dir[1][3:])

    if img.size[0] > img.size[1]:
        new_img = img.crop( (
            (((img.size[0])-120)/2),
            0,
            120+(((img.size[0])-120)/2),
            120
        ) )
    else:
        new_img = img.crop( (
            0,
            (((img.size[1])-120)/2),
            120,
            120+(((img.size[1])-120)/2)
        ) )

    new_img.save(file_dir[0]+'/ico'+file_dir[1][3:])

Based in @tomvon, I finished using the following (pick your case):

a) Resizing height (I know the new width, so I need the new height)

new_width  = 680
new_height = new_width * height / width 

b) Resizing width (I know the new height, so I need the new width)

new_height = 680
new_width  = new_height * width / height

Then just:

img = img.resize((new_width, new_height), Image.LANCZOS)

from PIL import Image
from resizeimage import resizeimage

def resize_file(in_file, out_file, size):
    with open(in_file) as fd:
        image = resizeimage.resize_thumbnail(Image.open(fd), size)
    image.save(out_file)
    image.close()

resize_file('foo.tif', 'foo_small.jpg', (256, 256))

I use this library:

pip install python-resize-image

from PIL import Image

img = Image.open('/your image path/image.jpg') # image extension *.png,*.jpg
new_width  = 200
new_height = 300
img = img.resize((new_width, new_height), Image.ANTIALIAS)
img.save('output image name.png') # format may what you want *.png, *jpg, *.gif

Just updating this question with a more modern wrapper
This library wraps Pillow (a fork of PIL)
https://pypi.org/project/python-resize-image/

Allowing you to do something like this :-

from PIL import Image
from resizeimage import resizeimage

fd_img = open('test-image.jpeg', 'r')
img = Image.open(fd_img)
img = resizeimage.resize_width(img, 200)
img.save('test-image-width.jpeg', img.format)
fd_img.close()

Heaps more examples in the above link.

A simple method for keeping constrained ratios and passing a max width / height. Not the prettiest but gets the job done and is easy to understand:

def resize(img_path, max_px_size, output_folder):
    with Image.open(img_path) as img:
        width_0, height_0 = img.size
        out_f_name = os.path.split(img_path)[-1]
        out_f_path = os.path.join(output_folder, out_f_name)

        if max((width_0, height_0)) <= max_px_size:
            print('writing {} to disk (no change from original)'.format(out_f_path))
            img.save(out_f_path)
            return

        if width_0 > height_0:
            wpercent = max_px_size / float(width_0)
            hsize = int(float(height_0) * float(wpercent))
            img = img.resize((max_px_size, hsize), Image.ANTIALIAS)
            print('writing {} to disk'.format(out_f_path))
            img.save(out_f_path)
            return

        if width_0 < height_0:
            hpercent = max_px_size / float(height_0)
            wsize = int(float(width_0) * float(hpercent))
            img = img.resize((max_px_size, wsize), Image.ANTIALIAS)
            print('writing {} to disk'.format(out_f_path))
            img.save(out_f_path)
            return

Here’s a python script that uses this function to run batch image resizing.

If you don’t want / don’t have a need to open image with Pillow, use this:

from PIL import Image

new_img_arr = numpy.array(Image.fromarray(img_arr).resize((new_width, new_height), Image.ANTIALIAS))

I was trying to resize some images for a slideshow video and because of that, I wanted not just one max dimension, but a max width and a max height (the size of the video frame).
And there was always the possibility of a portrait video…
The Image.thumbnail method was promising, but I could not make it upscale a smaller image.

So after I couldn’t find an obvious way to do that here (or at some other places), I wrote this function and put it here for the ones to come:

from PIL import Image

def get_resized_img(img_path, video_size):
    img = Image.open(img_path)
    width, height = video_size  # these are the MAX dimensions
    video_ratio = width / height
    img_ratio = img.size[0] / img.size[1]
    if video_ratio >= 1:  # the video is wide
        if img_ratio <= video_ratio:  # image is not wide enough
            width_new = int(height * img_ratio)
            size_new = width_new, height
        else:  # image is wider than video
            height_new = int(width / img_ratio)
            size_new = width, height_new
    else:  # the video is tall
        if img_ratio >= video_ratio:  # image is not tall enough
            height_new = int(width / img_ratio)
            size_new = width, height_new
        else:  # image is taller than video
            width_new = int(height * img_ratio)
            size_new = width_new, height
    return img.resize(size_new, resample=Image.LANCZOS)

I resizeed the image in such a way and it’s working very well

from io import BytesIO
from django.core.files.uploadedfile import InMemoryUploadedFile
import os, sys
from PIL import Image


def imageResize(image):
    outputIoStream = BytesIO()
    imageTemproaryResized = imageTemproary.resize( (1920,1080), Image.ANTIALIAS) 
    imageTemproaryResized.save(outputIoStream , format='PNG', quality='10') 
    outputIoStream.seek(0)
    uploadedImage = InMemoryUploadedFile(outputIoStream,'ImageField', "%s.jpg" % image.name.split('.')[0], 'image/jpeg', sys.getsizeof(outputIoStream), None)

    ## For upload local folder
    fs = FileSystemStorage()
    filename = fs.save(uploadedImage.name, uploadedImage)

Have updated the answer above by “tomvon”

from PIL import Image

img = Image.open(image_path)

width, height = img.size[:2]

if height > width:
    baseheight = 64
    hpercent = (baseheight/float(img.size[1]))
    wsize = int((float(img.size[0])*float(hpercent)))
    img = img.resize((wsize, baseheight), Image.ANTIALIAS)
    img.save('resized.jpg')
else:
    basewidth = 64
    wpercent = (basewidth/float(img.size[0]))
    hsize = int((float(img.size[1])*float(wpercent)))
    img = img.resize((basewidth,hsize), Image.ANTIALIAS)
    img.save('resized.jpg')

Open your image file

from PIL import Image
im = Image.open("image.png")

Use PIL Image.resize(size, resample=0) method, where you substitute (width, height) of your image for the size 2-tuple.

This will display your image at original size:

display(im.resize((int(im.size[0]),int(im.size[1])), 0) )

This will display your image at 1/2 the size:

display(im.resize((int(im.size[0]/2),int(im.size[1]/2)), 0) )

This will display your image at 1/3 the size:

display(im.resize((int(im.size[0]/3),int(im.size[1]/3)), 0) )

This will display your image at 1/4 the size:

display(im.resize((int(im.size[0]/4),int(im.size[1]/4)), 0) )

etc etc

I will also add a version of the resize that keeps the aspect ratio fixed.
In this case, it will adjust the height to match the width of the new image, based on the initial aspect ratio, asp_rat, which is float (!).
But, to adjust the width to the height, instead, you just need to comment one line and uncomment the other in the else loop. You will see, where.

You do not need the semicolons (;), I keep them just to remind myself of syntax of languages I use more often.

from PIL import Image

img_path = "filename.png";
img = Image.open(img_path);     # puts our image to the buffer of the PIL.Image object

width, height = img.size;
asp_rat = width/height;

# Enter new width (in pixels)
new_width = 50;

# Enter new height (in pixels)
new_height = 54;

new_rat = new_width/new_height;

if (new_rat == asp_rat):
    img = img.resize((new_width, new_height), Image.ANTIALIAS); 

# adjusts the height to match the width
# NOTE: if you want to adjust the width to the height, instead -> 
# uncomment the second line (new_width) and comment the first one (new_height)
else:
    new_height = round(new_width / asp_rat);
    #new_width = round(new_height * asp_rat);
    img = img.resize((new_width, new_height), Image.ANTIALIAS);

# usage: resize((x,y), resample)
# resample filter -> PIL.Image.BILINEAR, PIL.Image.NEAREST (default), PIL.Image.BICUBIC, etc..
# https://pillow.readthedocs.io/en/3.1.x/reference/Image.html#PIL.Image.Image.resize

# Enter the name under which you would like to save the new image
img.save("outputname.png");

And, it is done. I tried to document it as much as I can, so it is clear.

I hope it might be helpful to someone out there!

import cv2
from skimage import data 
import matplotlib.pyplot as plt
from skimage.util import img_as_ubyte
from skimage import io
filename='abc.png'
image=plt.imread(filename)
im=cv2.imread('abc.png')
print(im.shape)
im.resize(300,300)
print(im.shape)
plt.imshow(image)

The following script creates nice thumbnails of all JPEG images preserving aspect ratios with 128×128 max resolution.

from PIL import Image
img = Image.open("D:\Pictures\John.jpg")
img.thumbnail((680,680))
img.save("D:\Pictures\John_resize.jpg")

You can combine PIL’s Image.thumbnail with sys.maxsize if your resize limit is only on one dimension (width or height).

For instance, if you want to resize an image so that its height is no more than 100px, while keeping aspect ratio, you can do something like this:

import sys
from PIL import Image

image.thumbnail([sys.maxsize, 100], Image.ANTIALIAS)

Keep in mind that Image.thumbnail will resize the image in place, which is different from Image.resize that instead returns the resized image without changing the original one.

EDIT: Image.ANTIALIAS raises a deprecation warning, and will be removed in PIL 10 (July 2023). Instead, you should use Resampling.LANCZOS:

import sys
from PIL import Image
from PIL.Image import Resampling

image.thumbnail([sys.maxsize, 100], Resampling.LANCZOS)

The simplest way that worked for me

image = image.resize((image.width*2, image.height*2), Image.ANTIALIAS)

Example

from PIL import Image, ImageGrab
image = ImageGrab.grab(bbox=(0,0,400,600)) #take screenshot
image = image.resize((image.width*2, image.height*2), Image.ANTIALIAS)
image.save('Screen.png')

To make the new image half the width and half the height of the original image, Use below code :

  from PIL import Image
  im = Image.open("image.jpg")
  resized_im = im.resize((round(im.size[0]*0.5), round(im.size[1]*0.5)))
    
  #Save the cropped image
  resized_im.save('resizedimage.jpg')

To resize with fixed width with ration:

from PIL import Image
new_width = 300
im = Image.open("img/7.jpeg")
concat = int(new_width/float(im.size[0]))
size = int((float(im.size[1])*float(concat)))
resized_im = im.resize((new_width,size), Image.ANTIALIAS)
#Save the cropped image
resized_im.save('resizedimage.jpg')

# Importing Image class from PIL module
from PIL import Image

# Opens a image in RGB mode
im = Image.open(r"C:UsersSystem-PcDesktopybear.jpg")

# Size of the image in pixels (size of original image)
# (This is not mandatory)
width, height = im.size

# Setting the points for cropped image
left = 4
top = height / 5
right = 154
bottom = 3 * height / 5

# Cropped image of above dimension
# (It will not change original image)
im1 = im.crop((left, top, right, bottom))
newsize = (300, 300)
im1 = im1.resize(newsize)
# Shows the image in image viewer
im1.show()

######get resize coordinate after resize the image using this function#####
def scale_img_pixel(points,original_dim,resize_dim):
        multi_list = [points]
        new_point_list = []
        multi_list_point = []
        for point in multi_list:
            multi_list_point.append([point[0],point[1]])
            multi_list_point.append([point[2],point[3]])
        for lsingle_point in multi_list_point:
            x1 = int((lsingle_point[0] * (resize_dim[0] / original_dim[0])))
            y1 = int((lsingle_point[1] * (resize_dim[1] / original_dim[1])))
            new_point_list.append(x1)
            new_point_list.append(y1)
            
        return new_point_list
    
    
    points = [774,265,909,409]
    original_dim = (1237,1036)
    resize_dim = (640,480)
    result = scale_img_pixel(points,original_dim,resize_dim)
    print("result: ", result)