How to visualize numpy.zeros array as an RGB image?

Question

I am working on Ubuntu using python 2.7 with OpenCV I am trying to display 4 images in the same window. Because opencv does not have any function for this I will try to make one. I have search in Stack Overflow and I found some complicated solutions but generated in C/C++ which I do not know.

So, my strategy is: I have 4 color BGR images. All 4 same size: (x,y) I will generate a numpy array of zeros 4 times the size of the original images (2x,2y) I will copy each images in the numpy zero array, but in different position, so they will look each image next to the other:

import cv2
import numpy as np

def ShowManyImages():




    img1 = cv2.imread('img1.png',1)
    img2 = cv2.imread('img2.png',1)
    img3 = cv2.imread('img3.png',1)
    img4 = cv2.imread('img4.png',1)

    x,y,_ = img1.shape
    print x, y   # show: 500,500
    img_final = np.zeros((x*2, y*2, 3), np.uint8)
    print img_final.shape   # show: 1000,1000,3
    img_final[0:500,0:500,:] = img1[:,:,:]

    cv2.imshow('final', img_final)


    cv2.waitKey()

    cv2.destroyAllWindows()



ShowManyImages()

However, the code does not show any image. I can't figure out why. Any ideas?

Note: to make the code shorter, I only show the code for the first img (img1)

Using numpy.bmat might be cleaner giving a numpy.matrix that would need conversion to array. — Divakar
– Divakar, Commented Apr 27, 2017 at 21:46
Thank you for the input. I did not know this option. I tried bmat but I got an error message: 'ValueError: shape too large to be a matrix.' I went to the documentation and I undersand is only for 2D arrays?'out : matrix Returns a matrix object, which is a specialized 2-D array. ' — daniel_hck
– daniel_hck, Commented Apr 27, 2017 at 21:57

Divakar · Accepted Answer · 2017-04-27 22:06:10Z

2

Here's an approach with few stack operations assuming a0,a1,a2,a3 as the four 3D (RGB) image arrays -

a01 = np.hstack((a0,a1))
a23 = np.hstack((a2,a3))
out = np.vstack((a01, a23))

Sample run -

In [245]: a0 = np.random.randint(0,9,(2,3,3))
     ...: a1 = np.random.randint(0,9,(2,3,3))
     ...: a2 = np.random.randint(0,9,(2,3,3))
     ...: a3 = np.random.randint(0,9,(2,3,3))
     ...: 

In [246]: a01 = np.hstack((a0,a1))
     ...: a23 = np.hstack((a2,a3))
     ...: out = np.vstack((a01, a23))
     ...: 

In [247]: out.shape
Out[247]: (4, 6, 3)

Thus, we would have them stacked like so -

a0 | a1
-------
a2 | a3

answered Apr 27, 2017 at 22:06

Divakar

222k19 gold badges273 silver badges374 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How to visualize numpy.zeros array as an RGB image?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related