CODEWITHSUNDEEP
python-3.xnumpyopencvimage-manipulation
Annie Wojciek
Annie Wojciek

Reputation: 43

How to resize the window obtained from cv2.imshow()?

I started learning OpenCV today and I wrote a short code to upload (I don't know, if it's the right term) a random image:

enter image description here

It works fine, and I can open the image, but what I get is a big window and I can't see the full image unless I scroll it:

enter image description here

So, I'd like to know a way that I could see the whole image pretty and fine in a shorter window.

Upvotes: 4

Views: 17016

Answers (3)

Semnodime
Semnodime

Reputation: 2003

When cv2.imshow() creates a new window, it passes the flag cv::WINDOW_AUTOSIZE, preventing resize.

To resize an already existent window, you have to disable this flag:

import cv2
import numpy as np

if __name__ == '__main__':
    cv2.imshow('image', np.ones((128, 128, 3)))
    cv2.setWindowProperty('image', 1, cv2.WINDOW_NORMAL)
    cv2.waitKey()
    cv2.resizeWindow('image', 512, 512)
    cv2.waitKey()
    exit()

Before
enter image description here
After
enter image description here

Upvotes: 2

HansHirse
HansHirse

Reputation: 18925

The actual "problem" comes from imshow itself, and is the following:

If the window was not created before this function, it is assumed creating a window with cv::WINDOW_AUTOSIZE.

Looking at the corresponding description at the WindowFlags documentation page, we get:

the user cannot resize the window, the size is constrainted by the image displayed.

So, to work around this, you must set up the window manually using namedWindow, and then resize it accordingly using resizeWindow.

Let's see this code snippet:

import cv2

# Read image
image = cv2.imread('path/to/your/image.png')

# Window from plain imshow() command
cv2.imshow('Window from plain imshow()', image)

# Custom window
cv2.namedWindow('custom window', cv2.WINDOW_KEEPRATIO)
cv2.imshow('custom window', image)
cv2.resizeWindow('custom window', 200, 200)

cv2.waitKey(0)
cv2.destroyAllWindows()

An exemplary output would look like this (original image size is [400, 400]):

Output

Using cv2.WINDOW_KEEPRATIO, the image is always fitted to the window, and you can resize the window manually, if you want.

Hope that helps!

Upvotes: 18

flamelite
flamelite

Reputation: 2854

You can resize the image keeping the aspect ratio same and display it. 

#Display image
def display(img, frameName="OpenCV Image"):
    h, w = img.shape[0:2]
    neww = 800
    newh = int(neww*(h/w))
    img = cv2.resize(img, (neww, newh))
    cv2.imshow(frameName, img)
    cv2.waitKey(0)

Upvotes: 4

Related Questions